3116124, 8:08 PM capstone projectreport, HOUSE PRICE PREDICTION Structure: 1. Model Building with Dataset-1 2. Hypertuning Dataset-1 3. Summary - Dataset-1 4, Model Building with Dataset-2 5. Hypertuning Dataset-2 6. Summary - Dataset -2 7. Conclusion Note: Dataset - 1 = 22 features [price’, ‘room_bed’, room_bath’, living_measure’, ot_measure’, ‘ceil’, ‘coast’, ‘sight, ‘condition’, quality, ‘ceil_measure’ r_built, tiving_measure15,, lol_measure15,, furnished’, ‘otal_area’, ‘month_year, ‘city’, has_basement,, ‘HouseLandRatio’, ‘has_renovated'] Dataset - 2 = 31 features (important features after imputing dummy and analyzing different models) [price’, room_bed, ‘room_bath, tiving_measure’, ot_measure’, ‘ceil, ‘sight, ‘condition’, ‘ceil_measure’, ‘basement, 'yr_buitt, 'yr_renovated’, ‘zipcode’, at’ ong’, tiving_measure15’, tot_measure15, total_area’, ‘coast_1, ‘quality 3’ ‘quality 4, ‘quality_5’ ‘quality_6; ‘quality_7, quality_8', quality 9°, ‘quality 10, ‘quality_11’ ‘quality_12', ‘quality_13", furnished_1" Model building Let's build the model and see their performances Linear Regression (with Ridge and Lasso) In [135]: | #importing the necessary Libraries from sklearn.linear_model import LinearRegression from sklearn.linear_model import Ridge from sklearn.linear_model import Lasso from sklearn import metrics from sklearn.metrics import r2_score, mean_squared_error, mean_absolute_error {ie:1IC:1Users/SAUMYA SINGH/DocumenisiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Mtn we3116124, 8:08 PM In [136]: capstone projectreport, LR1 = LinearRegression() LR1.#it(X_train, y_train) tpredicting result over test data y_LR1_predtr= LRi.predict (x_train) y_LR1_predvl= LRi.predict (x_val) LR1.coef_ out[136]: array([ 4.65340600e+01, -2,04412513e+03, 1.71493571e+02, -2,49409439e+04, -6,13484472e+04, 1.25982999e+04, 9.03633165e+04, 1.12536191e+05, 1.64433901¢+05, 1.74737226e+05, -7.18864612e-09, 5.71678128e+04, 4,57278795e+04, 1.30265390e+05, -1.50728574e105, -3.31240159¢+05,, 4,729711260+05, 1.76625819e+85, -1,34952467e+04, 2.33952638e+05, 1.32976614e+85, 1.73301885e+85, 2.80a14383e+04, -2.60438290e+01, 5.35189660e+01, -1.52366392e+04, -3. 384286500104, -5.01063903e+04, 8.20853745e+04, 1.00072705e+05, 1.13322504e+05, 2.85527951e+05, 9.19752797e+05, 1.64060215e+04, 7.56908538e+04, 1.17418144e405, 1.58016139e+05, -1.30295440e105, -2.05150668e+05, 2.94909682e+05, 1.09165616e+05, 1.66521627¢+@5, 4,074064480+04, 2.32526660e+05, 1.04083333e+05, 3.11287808e+04]) 4.15085952e+01, -1,85866571e+00, -1.71559917e+02, -7.05629386e+04, -1.54205825e+05, 9.07899656e+04, 1.08362140e+05, 1.30547366e105, 1.71382012e+05, 1,55294652e105, 1.53741629e+04, 2.60659481e+05, 2.55845105e+05, 1.95914534e+05, ~5.38413267e104, 3,46837714e104, 1,32458120e105, 1.81816493e104, 1.21199592¢+05, 8.02211265¢+05, 6.01053338¢+04, 8.65668543e+04, {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm 5.02547332e+00, 2.04902213e+01, -7,41190980e+03, -1.39570745e105, -2, 246568300105, 2.27686059e+05, 1.12686599e+05, 1.77676262e+05, 1,61500051e+05, 2.95336027e+05, 5.55177883e+04, 4,01215839e+04, 9.85063804e+04, -1.522253010+05, 1.41195146e+04, 9.74678182e+05, 1.23710826e+05, 1.72163167e+05, 1.50478273¢+04, 4.196843840+05, 1.61179525e+05, 1.56470485e+25, 2783716724, 8:06 PM In [137]: out [137]: capstone projectreport, f#odel score and Deduction for each Model in a DataFrame LR1_trscore=r2_score(y_train,y_LR1_predtr) LR1_trRMSE=np.sqrt(mean_squared_error(y train, y_LR1_predtr)) LR1_teMSE=mean_squared_error(y train, y_LRi_predtr) LR1_teMAE=mean_absolute_error(y train, y_LRi_predtr) LR1_viscore=r2_score(y_val,y_LR1_predvi) LR1_vIRMSE=np.sqrt(mean_squared_error(y val, y_LR1_predvl)) LR1_vINSE-mean_squared_error(y val, y_LR1_predvl) LR1_v1MAE=mean_absolute_error(y_val, y_LR1_predvl) Compa_df=pd.DataFrame({'Method' :['Linear Reg Model1'], "Val Score’ :LR1_vlscor e,'RMSE_v1": LR1_vIRMSE, 'MSE_v1': LR1_vIMSE, 'MAE_vi': LR1_vIMAE, ‘train Scor e':LR1_trscore,'RMSE_tr': LR1_trRMSE, 'MSE_tr': LR1_trMSE, ‘MAE tr’: LR1_trMA E}) #Compa_df = Compa_df[['Method’, ‘Test Score’, 'RMSE', 'MSE', ‘MAE’]] Compa_d# Method Nal RMSE_vI MSE_vi waz tain RMSE_t od Score Ml Ml “Score tr Linear © Reg 0.718749 137733.698415 1.897057e+10 93994.455301 0.790112 132958.367261 1.7677 Modett The linear regression model performed with scores 0.73 & .72 in training data set and validation data set respectively {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm size3116124, 8:08 PM capstone projectreport, In [138]: sns.set(style="darkgrid", color_codes-True) with sns.axes_style(“white"): sns.jointplot(x=y_val, _LR1_predvl, kind="reg", colo’ 2000000 ‘800000 ‘000000 00000 ‘© 500000 1090000 1500000 2000000 2500000 rice Lasso model {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm 4n783116124, 8:08 PM In [139]: Lasso1 = Lasso(alpha=1) capstone projectreport, Lasso1.fit(X_train, y_train) apredicting result over test data y_Lassol_predtr= Lasso1.predict(X_train) y_Lassol_predvi= Lasso1. predict (X_val) Lassol.coef_ cause precision problens. ConvergenceWarning) out[139}: array([ 9.65902931e+01, -2.04163675e+03, 1.73465396e+02, 1.64885732e+04, -1.65210854e+04, -9,13241618e+04, -1.64860326e+04, 5.60351070e+03, 5.73457320e+04, 6.65775444e+04, 2.20000000e+00, 5. 707683440404, 4,57348433¢104,, 1.25980055e+05, =2.00695519e+05,, 5.98641869e+04, 3.18911457e+04, 1.75754198e+85, -1,38642363e+04, 2.33538555e+05, 1.32466072e+85, 1.72962017e+85, 2.8aa9saade+ad, -3.93953142e+00, 5.35553554e+01, 2.58275091e+04, 7.61896026€+03, -4,12121513e+03, -2.48925268e+04, -6.79525494e+03, 6.32472345e+03, 1.78131982e+05, 8.00797041e+05, 1.64275884e+04, 7.272843800+04, 1.17367493e105, 1.53729964e405, -1,80295562e+05, 1.85986501¢+05, 2.944658260+05, 1.0862595ee+05, 1.66073186¢+05, 4,03015196e+04, 2.32091780¢+05, 1.03839270e+05, 3.11881238e+04]) 1.35529095e101, -1.85807486e+00, 4.11967331e+04, -2.86912815e+04, -1.00992716e+05, -1,60169381e+04, 1.46220219e+03, 2.34670403e+04, 6.38573909e+04, 4,05485728e+04, 1.5346a259e+04, 2.60559411¢+05, 2.55764846e105, 1,91681487e+05, -1.03865243e105, 4,26066854e+05, 1.31572941¢405, 1. 76695808e+04, 1.20696485¢105, 8.01308223e+05, 5.96809785¢104, 8.54647044e+04, {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm : \ProgramData\Anaconda3\1ib\site-packages\sklearn\linear_model\coordinate_de scent.py:492: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may -2,29061776e+01, -1.60949858e+00, 3.40052776e+04, -9.69275851e+04, -1.71298527e+05, 1.18516529e+05, 5.75895038e+03, 7.06751057e+04, 5.27207711e+04, 1,84114085e+05, 5.53276759e+04, 3.99774418¢+04, 9.44643505e+04, -1.980774200+05, -3.58821013e104, 1,35396787e+06, 1.231997460+05, 1.70928679e+05, 1.45879439e+04, 4,19176853e+05, 1.60725157e+05, 1.55943713e+25, 51783116124, 8:08 PM capstone projectreport, In [140]: #Model score and Deduction for each Model in a Datafrane Lassol_trscore=r2_score(y_train,y_Lassol_predtr) Lassol_trRMsE=np.sqrt(mean_squared_error(y train, y_Lassol_predtr)) Lassol_trusé=mean_squared_error(y train, y_Lassol_predtr) Lassol_trMAé=mean_absolute_error(y train, y_Lassol_predtr) Lassol_viscore=r2_score(y_val, y_Lasso1_predvl) Lasso1_v1RMSE=np.sqrt(mean_squared_error(y_val, y_Lasso1_predvl)) Lassol_vIMSE=mean_squared_error(y_val, y_Lassol_predvl) Lassol_vIMAE=mean_absolute_error(y_val, y_Lassoi_predvi) Lasso1_df=pd.DataFrame({'Method' :['Linear-Reg Lasso1'], ‘Val Score’ :Lasso1_visc ore, 'RNSE_vl': Lassol_vIRMSE, "MSE_v1': Lasso1_vINSE, 'MAE_vl': Lassoi_v1MA E,'train Score’ :Lassol_trscore,'RMSE_tr’: Lassol_trRMSE, 'MSE_tr': Lasso1_trMs E, 'MAE_tr': Lassol_trMAE}) Compa_d# = pd.concat([Conpa_df, Lassol_df]) Compa_dt out[140]: Val train Method val RMSE_vI MSE_vi mae tain RMSE_tr Linear © Reg 0.718749 137733.698415 1.897057e+10 99994.455301 0.730112 132958367261 1.7671 Model! Linear 0 Rog 0.719117 137643.639712 1.894577e+10 93939441186 0.730092 132963.180396 1.767! Lasso! ‘The lasso linear regression model performed with scores 0.73 & .72 in training data set and validation data set respectively. The coefficeints of 1 variable in lasso model is almost 0’, signifying that the variable with 'O" coefficient can be dropped. {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm 5783116124, 8:08 PM capstone projectreport, In [141]: sns.set(style="darkgrid", color_codes-True) with sns.axes_style(“white"): sns.jointplot(x=y_val, | Lassol_predvl, kind="reg", colo 2000000 ‘800000 ‘000000 00000 ‘© 500000 1090000 1500000 2000000 2500000 rice Ridge model {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm mre3116124, 8:08 PM In [142]: Ridge1 = Ridge(alpha-o. Ridgel.fit(X_train, y_t apredicting result over y_Ridgel_predtr= Ridge. y_Ridgel_predvl= Ridge1. Ridget.coef_ out[142]: array([ 4.66834622e+01, -2.04329070e+03, 1.99149390e+02, 3.84394798e+04, 2.07847666e+03, -1.24763131e+05, -6,18022272e+04, -4,00644775e+04, 1.13660948e+04, 2.07235857e+04, 2.e0000020e+00, 5.68361232e+04, 4,584989300+04,, 1.12963518e+05, -1.40442366¢+05, -2.64929900+05, 4,12010220e+05, 1.69604765¢+05, -1.74612788¢+04, 2.29612234e+85, 1.28389461e+05, 1.69467429e+05, 2.85123132e+04, capstone projectreport, 5) rain) test data -predict(x_train) spredict (X_val) -2.60918244e+01, 5 .40305072e+01, 4.75922037¢+04, 2.95688617e+04, 1.15197429e+04, -6.99157975e+04, -5.21975799e+04, -3.93829165e+04, 1,31445711e+05, 5 .06032623e+05, 1.62595610e+04, 6.73075467e+04, 1.16570206e+05, 1.40760828e+05, -1.20195568e+05, -1,34329618e+05, 2.90368008¢+05, 1.04432882c+@5, 1.62055462e+@5, 3.67503018¢+04, 2.28151004e+05, 9.98907258e+04, 3.11686377e+04]) 4.15530007e+01, -1.83894732e+00, 6.34342640e+04, -6.95633001e+03, -6.08136975e+04, -6.12127626e+04, -4,39150711e+04, -2.22301737e+04, 1, 794g0665e+04, 1.96912226e+03, 1,52860502e+04, 2,58145115e+05, 2,54669644e+05, 1.78571191¢105, 4, 38816320e104, 1,11752710e105, 1. 26135853e105, 1.42752316¢+04, 1.16592568¢105, 7.71949916e+05, 5.60163760e104, 8.00564608e+04, {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm 5.13138900e+00, 2.05899911e+01, 5.62050467e+04, -7.29503951e+04, -1.07506532e+05, 684637990104, -3.96547935e104, 2.63190006e+04, 7.08429226e+03, 1,26354803e+05, 5.48436035¢+04, 3.96081664e+04, 8.12334931¢+04, 1. 31486307e105, 2,39956530e+04, 6.95080119e+05, 1,18985675e+05, 1.63147681¢+05, 1.10585172e+04, 4.13369607e+05, 1.56620270e+25, 1.51440054e+05, a73116124, 8:08 PM capstone projectreport, In [143]: | #Model score and Deduction for each Model in a DataFrame Ridge1_trscore=r2_score(y_train,y Ridge1_predtr) Ridge1_trRMSE=np. sqrt (mean_squared_error(y train, y Ridge1_predtr)) Ridgel_trMSE-mean_squared_error(y train, y Ridgel_predtr) Ridgel_trMAE-mean_absolute_error(y train, y_Ridgel_predtr) Ridge1_viscore=r2_score(y_val,y_Ridgel_predv1) Ridge1_vIRMSE=np. sqrt (mean_squared_error(y_val, y_Ridge1_predvl)) Ridge1_vIMSE=mean_squared_error(y_val, y_Ridge1_predvl) Ridge1_vIMAE=mean_absolute_error(y_val, y_Ridgel_predv1) Ridge1_df=pd.DataFrame({ ‘Method’ :['Linear-Reg Ridge1'], 'Val Score’ :Ridget_visc ore, 'RNSE_vl': Ridgel_vIRMSE, "MSE_v1': Ridge1_vINSE, 'MAE_vl': Ridge1_v1MA E,'train Score’ :Ridgel_trscore,'RMSE_tr’: Ridgel_trRMSE, 'MSE_tr': Ridgel_trMs E, 'MAE_tr': Ridge1_trMAE)) Compa_d# = pd.concat([Compa_df, Ridge1_df]) Compa_dt out [143]: Val train Method val RMSE_vI MSE_vi mae tain RMSE_tr Linear 0 Reg 0.718749 137733.698415 1.897057e+10 99994.455301 0.730112 132958367261 1.7671 Model! Linear 0 Rog 0.719117 137643.639712 1.8945770+10 93939.441186 0.730092 132963.180396 1.767! Lasso! Linear- 0 Reg 0.718929 137689,597398 1.895843e+10 93992.809617 0.729789 133037.735155 1.769% Ridge? The Ridge linear regression model performed with scores 0.73 & .72 in training data set and validation data set respectively. The coefficeints of variables in ridge model are all non-zero, indicating that non of the variables can be dropped. {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm er7e3116124, 8:08 PM capstone projectreport, In [144]: sns.set(style="darkgrid", color_codes-True) with sns.axes_style(“white"): sns. jointplot(x=y_val, | Ridgel_predvl, kind="reg", color 2000000 ‘800000 ‘000000 00000 ‘© 500000 1090000 1500000 2000000 2500000 rice In summary, Linear models have performed almost with similar results in both regularized model and non- regularized models KNN Regressor In [145]: | from sklearn.neighbors import KNeighborsRegressor In [146]: knni = KNeighborsRegressor(n_neighbors=4,weights="distance') knn1.#it(X_train, y_train) fpredicting result over test data y_knn1_predtr= knn1.predict(X_train) y_knnd_predvl= knni. predict (Xx_val) {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm sone3116124, 8:08 PM capstone projectreport, In [147]: | #Model score and Deduction for each Model in a DataFrame knn4_trscore=r2_score(y_train,y_knn1_predtr) knnd_trRMSE=np.sqrt(mean_squared_error(y train, y_knn1_predtr)) knn1_trMSE=mean_squared_error(y_train, y_knni_predtr) knn1_trMAg-mean_absolute_error(y train, y_knni_predtr) knn_viscore=r2_score(y_val,y_knnt_predvi) knn1_v1RMSE=np. sqrt(mean_squared_error(y_val, y_knn1_predvl)) knt_vIMSé-nean_squared_error(y val, y_knni_predvl) kemt_vIMA-nean_absolute_error (y_val, y_knni_predvi) knn1_df=pd.DataFrame({*Method' :['knn1'], "Val Score’ :knni_viscore, 'RMSE_v1": kn na_vIRMSE, 'MSE_v1': knni_vIMSE, 'MAE_vl': knn1_vIMAE, ‘train Score" :knni_trsco re, 'RMSE_tr': knn1_trRMSE, 'MSE_tr': knn1_trMSE, ‘MAE tr’: knnd_trMAE}) Compa_d# = pd.concat([Compa_df, knn1_dt]) conpa_df out[147}: Method X!RMSEM = MSE MAEM gS Sete Cinear 0 "Reg oriera 19773309815 19070sTeet0 99004455001 a7a0rt2 12058.967261 1.76 Model Liner. 0 Rey O:719117 ta7e4aea6712 1e4s7Tev10 sasseaertes 0.720092 192860180090 1.761 Lasso’ Lincer- 0 eg 0718929 137680557208 1s9se4ser10 99992808617 0.729789 199007725155 1.76% Rege! 0 knnt 0.425008 196935.451160 3.878357e+10 138494.983286 0.998628 9480.192071 8.981 ‘Though KNN regressor performed well in training set, the performance score in validation set is very less. This shows that the model is overfited in training set Support vector regressor In [148]: from sklearn.svm import SVR In [149]: SVRL = SVR(gamma='auto", Sve1.fit(X_train, y_train) 12.2, epsilor .2,kernel='rof') y_SVR1_predti y_SVR1_predv: SVR1.predict(x_train) SVR1.predict(X_val) {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm ane3116124, 8:08 PM capstone projectreport, In [150]: #Model score and Deduction for each Model in a Dataframe SvR1_trscore=r2_score(y_train,y SVR1_predtr) SVR1_trRMSE=np.sqrt(mean_squared_error(y train, y_SVR1_predtr)) SVR1_trMSE=mean_squared_error(y_train, y_SVR1_predtr) SVR1_trMAE-mean_absolute_error(y train, y_SVR1_predtr) SVRI_viscore=r2_score(y_val,y_SVR1_predv1) ‘SVR1_v1RMSE=np. sqrt(mean_squared_error(y_val, y_SVR1_predv1l)) SVR1_vIMSE-nean_squared_error(y val, y_SVRi_predvl) SVRI1_vIMAE-nean_absolute_error(y_val, y_SVRi_predvl) SvR1_df=pd.DataFrame({ "Method! :['SVR1'], "Val Score’ :SVR1_vIscore, "RMSE_vl": SV R1_VvIRMSE, 'MSE_vl': SVRI_VIMSE, 'MAE_vl': SVR1_VIMAE, ‘train Score’ :SVR1_trsco re, 'RMSE_tr': SVRI_trRMSE, 'MSE_tr': SVR1_trMSE, ‘MAE tr’: SVR1_trMAE}) Compa_d# = pd.concat([Compa_df, SvR1_dt]) Conpa_af out (150): Method ValScore -RMSE_v| = MSE_yl mac U3 RMSE Linear 0 Reg 0.718749 137735.698415 1.897057e+10 99904.455301 0.730112 132958367261 1.71 Model! Linear. 0 Reg 0719117 197643.690712 1e94577e+10 93099.441188 0.73002 132960.180395 1.71 Lasso! Linear. 0 "Reg 0.718829 197689.597308 1e95843er10 99992.809617 0.729709 199097.795155 1.71 Ridge 0 knnt 0.425008 196936.451160 3.8783570+10 198494.383286 0.998628 9480.192071 8.91 0 SVRt -0,055489 266820.956555 7.1193420+10 183639,593215 -0,046405 261802341726 6.8! The above negative scores in SVR model is due to non-learning of the model in the training set which results in non-performance in validation set {ie:1IC:1Users/SAUMYA SINGH/DocumenisiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Mtn rane3116124, 8:08 PM capstone projectreport, In [151]: SVR2 = SVR(ganma=" auto" ,C-0.1,kernel=" linear’) SvR2.fit(X_train, y_train) y_SVR2_predt y_SVR2_predv. ‘SVR2. predict (X_train) SVR2. predict (X_val) #fodel score and Deduction for each Model in a DataFrame SvR2_trscore-r2_score(y_train,y_SVR2_predtr) SVR2_trRMSE=np. sqrt(mean_squared_error(y train, y_SVR2_predtr)) SVR2_trMSE=mean_squared_error(y train, ySVR2_predtr) SVR2_trMAE=mean_absolute_error(y_train, y_SVR2_predtr) ‘SVR2_vlscore=r2_score(y_val,y_SVR2_predvl) ‘SVR2_VIRMSE=np. sqrt (mean_squared_error(y_val, y_SVR2_predvl)) ‘SVR2_vIMSE=mean_squared_error(y_val, y_SVR2_predvl) ‘SVR2_VIMAE=mean_absolute_error(y_val, y_SVR2_predvl) ‘SvR2_d#=pd.DataFrame({*Method' :['SVR2'], ‘Val Score’ :SVR2_vIscore, "RMSF_v1': SV R2_vIRMSE, 'MSE_v1': SVR2_VIMSE, 'MAE_vl': SVR2_VINAE, ‘train Score’ :SVR2_trsco re, "RMSE_tr': SVR2_trRMSE, 'MSE_tr': SVR2_trMSE, 'MAE_tr': SVR2_trMAE}) Compa_df = pd.concat([Compa_df, SvR2_df]) Conpa_df Out[151]: Method ValScore = RMSEMI MSE yi mac 32 RMSE Ar Linear 0 "Reg 0718749 13733.698615 1.897057e+10 93004.455301 0.730112 132958.367261 1.7 Model! Linear 0 "Reg O71eH17 137643.690712 1.804577e+10 93899.461186 0730002 132969.180306 1.7 Lassot near 0 "Reg 0.718929 137689.597308 1.89584se+10 93092.809617 0.720789. 139087.796155 1.7 Ridge’ 0 kant 0.425008 196935.451160 3.878357e+10 138494.383286 0.998628 9480.192071 ai SVR1 -0,055489 266820.956555 7.119342e+10 183639,593218 -0,046405 261802.341726 6.8! 0 SVR2 0.458252 191157.623415 3.654124e+10 132876663665 0.454410 189041408746 3.5 ‘The SVR model with modified parameters has not performed well with just ~0.45 in both training and validation data sets De n Tree Regressor In [152]: from sklearn.tree import DecisionTreeRegressor {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm sane3116124, 8:08 PM In [153]: out [153]: capstone projectreport, DT1 = DecisiontreeRegressor() DT1.#it(X_train, y_train) y_DT1_predtr= DT1.predict(X_train) y_DT1_predvl= DT1.predict(X_val) #odel score and Deduction for each Model in a Dataframe DT1_trscore=r2_score(y_train,y_DT1_predtr) DT1_trRMSE=np. sqrt(mean_squared_error(y train, y DT1_predtr)) DT1_trMSE=mean_squared_error(y train, y_0T1_predtr) DT1_trWAE=mean_absolute_error(y train, y_0Ti_predtr) DT1_viscore=r2_score(y_val,y_DT1_predvl) DT1_vIRMSE=np.sqrt(mean_squared_error(y_val, y_DT1_predvl)) DT1_vIMSE=mean_squared_error(y_val, y_DT1_predvl) DT1_vIMAE-mean_absolute_error(y_val, y_DT1_predvl) DT1_d#=pd.DataFrame({ "Method! :['DT1'], 'Val Score’ :DT1_v1score, 'RMSE_vl': DT1_v ARMSE, 'MSE_vl': DT1_vINSE, 'MAE_vl': DT1_vINAE, train Score’ :DT1_trscore, ‘RMS E_tr': DTI_ErRMSE, 'MSE_te': DTI_trMSE, 'WAE_tr': DT1_trMAE}) Compa_df = pd.concat([Compa_df, DT1_df]) conpa_d¢ Motiod valSeore RMSE SEAR MRS Circo 0 Rog o7se749 ts77s3au6415 ‘eorosrert0 ss0e44ss501 o7GONI2 ‘seasesres1 17 wea Linea 0 hey o7tort7 tereessiort2 1eo4srrevio sseseaaries 0730002 132065180306 17 von! Linear okey 07%8820 te76e0sora6e ‘seaserto sooozene617 ar20Te9 so0s7. 705188 17 noe! 0 fon! 0426008 186505451160. saTa%eTer10 136404385286 O88 oHEO T2071 8 0 SRI 0086480. 266820986555 7tiozerto T80680805216 codes astenzeAi7z6 6.8 0 svme o4se252 totts7aza415 s6sétzHer10 T92876 658665 CASED 1HoDH1.8THB. 35 0 Dri ostaiss 175867376246 soesooterio 100001238551 oesece om8o 2071 8a ‘Above performance of initial Decision tree model shows overfit in training set with 0.99 score and low performance in validation set {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm sane3116124, 8:08 PM In [154]: out[154]: capstone projectreport, DT2 = DecisionTreeRegressor(max_depth=10,min_samples_leaf-5) DT2.Fit(X_train, y_train) y_DT2_predtr= DT2.predict(X_train) y_D12_predvl= D12.predict(X_val) #odel score and Deduction for each Model in a Dataframe DT2_trscore=r2_score(y_train,y_DT2_predtr) DT2_trRMSE=np..sqrt(mean_squared_error(y train, y DT2 predtr)) DT2_trMSEsmean_squared_error(y train, y_0T2_predtr) DT2_trWAE=mean_absolute_error(y train, y_0T2_predtr) DI2_vIscore=r2_score(y_val,y_DT2_predvl) DT2_vIRMSE=np.sqrt(mean_squared_error(y_val, y_DT2_predvl)) DT2_vIMSE=mean_squared_error(y_val, y_DT2_predvl) DT2_vIMAE-mean_absolute_error(y_val, y_DT2_predvl) D12_d#=pd.DataFrame({"Method' :['DT2'], 'Val Score’ :DT2_vIscore, 'RMSE_vl': DT2_v ARMSE, 'MSE_vl': DT2_vINSE, 'MAE_vl': DT2_vINAE, train Score’ :DT2_trscore, ‘RMS E_tr': DT2_ErRMSE, 'MSE_te': DT2_trMSE, 'WAE_tr': DT2_trMAE}) Compa_df = pd.concat([Compa_df, DT2_d#]) Compa_df Method Val Score RMSE_vi MSE_MI mac tain RMSE_tr Linear 0 Reg 0.78749 137733.698415 1.897057e+10 93994.455301 0.730112 132958.367261 1.71 Model! Linear- 0 Reg 0.719117 137643.639712 1.894577e+10 93999.441186 0.730082 132963.180396 1.71 Lasso! Linear- 0 Reg 0.718929 137689,597398 1,895843e+10 93992,809617 0.729789 133037.735155 1.71 Ridge? 0 kant 0.425008 196935.451160 3.878357e+10 138494.383286 0.998628 9480.192071 a9 0 SVRt -0,05489 266820.956555 7.119342e+10 183639,593215 -0,046405 261802341726 6.8! 0 SVR2 0.458252 191157.623415 3.654124e+10 132876.663665 0.454410 189041408746 3.5 0 DT 0.542495 175867376246 3.0859030+10 109891.238551 0.998628 9480.192071 8.91 0 DTZ 0.637513 156364,920550 2.4449990+10 102458,587308 0.794647 115977.718333 1.3 Above decision tree model with modified parameter has better performed on the training set and validation set compared to initial decision tree model.But overall decision tree has not performed well than linear regression models. {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm ssn783116124, 8:08 PM capstone projectreport, In [155]: sns.set(style="darkgrid", color_codes-True) with sns.axes_style(“white"): sns. jointplot(x=y_val, /_D12_predvl, kind="reg", colo’ 500000 000000 ‘800000 100000 00000 ‘© 500000 1090000 1500000 2000000 2500000 rice In summary, KNN regressor model and decision tree models have not performed well in comparison with linear regression models Ensemble techniques Boosting and Bag: In [156]: from sklearn.ensemble import GradientBoostingRegressor, BaggingRegressor {ie:1IC:1Users/SAUMYA SINGH/DocumenisiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Mtn see3116124, 8:08 PM In [157]: Out [157]: capstone projectreport, GB1-GradientBoostingRegressor(n_estimators - 208, learning rate = 0.1, random state=22) G61. Fit(X_train, y_train) y_G81_predtr= GB1.predict(X_train) y_G81_predvl-= GB1.predict(X_val) ‘Model score and Deduction for each Model in a DataFrane 681_trscore=r2_score(y_train,y_6B1_predtr) GB1_trRMSE=np. sqrt(mean_squared_error(y_train, y_GB1_predtr)) G81_trNSE-mean_squared_error(y train, y_GB1_predtr) G81_trlAE=nean_absolute_error(y_train, y_6B1_predtr) GB1_vlscore=r2_score(y_val,y_GB1_predvl) GB1_vIRMSE=np.sqrt(mean_squared_error(y_val, y_GB1_predvl)) GB1_vINSE-mean_squared_error(y_val, y_GB1_predvl) GB1_vINAE=mean_absolute_error(y_val, y_GB1_predvl) GB1_df=pd.DataFrame({"Method' :['GB1'], 'Val Score’ :GB1_v1score, ‘RMSE_vl': GB1_v 1RMSE, 'MSE_vl': GB1_vINSE, 'MAE_vl': GB1_vIMAE, ‘train Score’ :GB1_trscore, ‘RMS E_tr': GB1_trRMSE, 'MSE_te’: GB1_trMSE, 'MAE_tr': GB1_trMAE}) Compa_df = pd.concat([Compa_df, GB1_df]) conpa_a¢ Motied valseore RMSE SEM ARMY RS Crear 0 Rey o7se7s9 ts77s3se641s 1e0rosrert0 s9004456501 o7G0HI2 ‘seas8.e251 17 wea Linea 0 Rey o7toti7 toveessiort2 ‘eo4srrevio sssseaatiss oso. 2206018006 17 tone! Linear 0 “rey 0718828 187660807368 ‘oseeserto sooozene617 e720T89 s00H7. 705188 1.7 noe! 0 fon! 0426008 186855451160. s7E%67or10 136404385286 C8828 oHEO 1071 8. 0 SvRi 0086480. 266820986555 7tézerto 180600805016 nodes astenz4178 6.8 0 svme. 458252 tottsTaza415 sesétztert0 T328TSG55665 CASED HoDHLOSTHB. 35 0 OT! ote4s5 175667376246 soRseO%er10 10880 238551 a8 oHEE 12071 8. 0 DTz_ osersts 156964220550. 24440000r10 102458587908 O7OAGAT TSOTT. T1899 19: 0B! o7sziTt tet299H6208 1Asr2sTort0 52624310002 080821 sOHH0A 70598 14 Gradient boosting model has provided good scores in both training and validation sets {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm sme3116124, 8:08 PM In [158]: out[158]: capstone projectreport, 8GG1-BaggingRegressor(n_estimators=5@, oob_score= True,randon_state=14) 8GG1.Fit(X_train, y_train) y_BG61_predt y_BGG1_predv. BGG1.predict(x_train) BGG1.predict(x_val) #fodel score and Deduction for each Model in a DataFrame 8GG1_trscore-r2_score(y_train,y_8G61_predtr) 8GG1_trRMSE=np. sqrt(mean_squared_error(y train, y BGG1_predtr)) 8GG1_trMSE=mean_squared_error(y train, y_8GG1_predtr) 8GG1_trMAE=mean_absolute_error(y train, y_BG61_predtr) 8661_vlscore=r2_score(y_val,y_8GG1_predvl) 8GG1_VIRMSE=np. sqrt (mean_squared_error(y_val, y_BGG1_predvl)) BGG1_vIMSE=mean_squared_error(y_val, y_BGG1_predvl) 8GG1_VIMAE=mean_absolute_error(y_val, y_8GG1_predvl) BGG1_d#=pd.DataFrame({*Method' :['BGG1'], ‘Val Score’ :BGG1_vIscore, "RMSF_v1': BG G1_vIRMSE, 'MSE_v1':BGG1_VIMSE, 'MAE_vl': BGG1_vINAE, "train Score’ :BGG1_trscor fe, "RMSE_tr’: BGG1_trRMSE, 'MSE_tr': BGG1_trMSE, 'MAE_tr': BGG1_trMAE}) Compa_df = pd.concat([Compa_df, BGG1_d*]) Compa_dt Method Val Score RMSE_vi MSE_MI mac tain RMSE_tr Linear 0 Reg 0.78749 137733.698415 1.897057e+10 93994.455301 0.730112 132958.367261 1.71 Model! Linear- 0 Reg 0.719117 137643.639712 1894577010 93999.441186 0.730082 132963180396 1.71 Lasso! Linear- 0 Reg 0.718929 137689,597398 1,895843e+10 93992,809617 0.729789 133037.735155 1.71 Ridge? 0 kant 0.425008 196935451160 3.878357e+10 138494.383286 0.998628 9480.192071 ai 0 SVRt -0,055489 266820.956555 7.119342e+10 183639,593215 -0,046405 261802341726 6.8! 0 SVR2 0.458252 191157.623415 3.654124e+10 132876663665 0.454410 189041408746 3.5 0 DT 0.542495 175867376246 3.0859030+10 109891.238551 0.998628 9480192071 8.91 0 DTZ 0.637513 156364,920550 2.4449990+10 102458,587308 0.794647 115977.718333 1.3 0 © GB1_ 0.782471 121129.980228 1.467247e+10 82624.319932 0.820821 108334.766538. 1.1 0 BGGI 0.769319 124738.101557 1.555959e+10 80102,360544 0.966466 46867.181534 2.1! Bagging model also performed well in training and validation sets. There seems to be overfitting in training set. ‘We need to analyse further by hypertuning {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm sere3116124, 8:08 PM capstone projectreport, Random forest In [159 from sklearn.ensemble import RandonForestRegressor {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm sen7e3116124, 8:08 PM In [169]: out[160]: capstone projectreport, RF1-RandomForestRegressor() RF1.fit(X_train, y train) y_RF1_predtr: y_RF1_predvl.: RF1.predict(x_train) RF1.predict(Xx_val) #Model score and Deduction for each Model in a Dataframe RF1_trscore=r2_score(y_train,y_RF1_predtr) RF1_trRMSE=np..sqrt(mean_squared_error(y train, y_RF1_predtr)) RF1_trWSE=mean_squared_error(y train, y_RF1_predtr) RF1_trMAE=mean_absolute_error(y train, y_RF1_predtr) RF1_vIscore=r2_score(y_val,y_RF1_predvl) RF1_vIRMSE=np.sqrt(mean_squared_error(y_val, y_RF1_predvl)) RF1_vINSE-mean_squared_error(y_val, y_RF1_predvl) RF1_vINAE=mean_absolute_error(y_val, y_RF1_predvl) RF1_df=pd.DataFrame({"Method' :['RF1'], 'Val Score’ :RF1_vIscore, 'RMSF_vl': RF1_v IRMSE, 'MSE_vl':RF1_vIMSE, 'MAE_v1': RF1_vIMAE,'train Score’ :RF1_trscore, ‘RMSE _tr': RFI_tPRMSE, 'NSE_tr': RF1_trMSE, 'MAE_tr': RF1_trMAE}) Compa_df = pd.concat([Compa_df, RF1_df]) Compa_df Method Val Score RMSE_vi MSE_MI mac tain RMSE_tr Linear 0 Reg 0.78749 137733.698415 1.897057e+10 93994.455301 0.730112 132958.367261 1.71 Model! Linear- 0 Reg 0.719117 137643.639712 1.894577e+10 93999.441186 0.730082 132963.180396 1.71 Lasso! Linear- 0 Reg 0.718929 137689,597398 1,895843e+10 93992,809617 0.729789 133037.735155 1.71 Ridge? 0 kant 0.425008 196935.451160 3.878357e+10 138494.383286 0.998628 9480.192071 a.ai 0 SVRt -0,085489 266820.956555 7.119342e+10 183639,593215 -0,046405 261802341726 6.8! 0 SVR2 0.458252 191157.623415 3.654124e+10 132876.663665 0.454410 189041408746 3.5 0 -DTt 0.542495 175867376246 3.0859030+10 109891.238551 0.998628 9480192071 8.91 0 DTZ 0.637513 156364,920550 2,.4449990+10 102458,587308 0.794647 115977.718333 1.3 0 © GBt_ 0.782471 121129.980228 1.467247e+10 82624.319932 0.820821 108394.766538 1.1 0 BGGI 0.769319 124738,101557 1.555959e+10 80102,360544 0.986466 46867.181534 2.1! 0 —-RF1 0.754483 128686.871977 1.6560310+10 82901.717082 0.954362 54674.891629 2.91 Random forest model has performed well in training and validation set. There is scope of further analysis on this model {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm 201783716724, 8:06 PM capstone projectreport, Enseble models: in summary ensemble models have performed well on training and validation sets. These models will be selected for further analysis with hypertuning and feature selection In [161]: #feature importance rf_imp_feature_1-pd.Datarrame(RF1.feature_importances_, columns = [“Imp"], ind ex = X_val. columns) rf_imp_feature_1.sort_values(by="Inp" ,ascending-False) rf_imp_feature_1[‘Imp"] = rf_imp_feature_1[‘Imp'].map("{0:.5f}" format) r#_imp_feature_1=rf_imp_feature_1.sort_values(by="Imp" ,ascending-False) rf_imp_feature_1.Inp=rf_imp_feature_1.Inp.astype("float") rf_imp_feature_1[:30].plot.bar(figsize-(plotsizex, plotsizey)) 4First 26 features have an importance of 98.5% and first 30 have importance of 95.15 print("First 20 feature importance: \t", (rf_imp_feature_1 print("First 30 feature importance: \t", (rf_imp_feature_1 20]. sum())*180) 30] .sum())*180) First 20 feature importance: Imp 98.184 dtype: floates First 30 feature importance: Imp 95.098 dtype: floates = i aes oa aro im Lae A a aac nEEeT ees si gee ER ETAN EE DE 1404 PRLEF EEE RFR GES LR eC ZER TG Yigi esapeipebbaeaspgi Pappas agit “fe pss B Fe Sa | @ § a8 z g 3 aé ‘Above are top 30 important features that account for 95% of variation in model. This need to be further analysed during hypertuning of the models for better scores {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm aire3116124, 8:08 PM capstone projectreport, Model performance Summary: Ensemble methods are performing better than linear models. Of all the ensemble models, Gradient boosting regressor is giving better R2 score. we identified top 30 features that are explaining the 95% variation in model(Random Forest). Will further hypertune the model to improve the model performance. Will further explore and evaluate the features while hyperturning the ensemble models {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm ante3116124, 8:08 PM Building Function/Pipeline for models In [162]: rf_imp_feature_1[:30] out [162]: furnished_4 yr_buitt living_measure living_measurets quality HouseLandRatio lot_measuret5 City Bellevue coll_measure quality. 8 total_area lot_measure City Seattle City_Kirkland City Federal Way ity_Kent City_ Mercer island sight 4 quality_7 basement City_Redmond coast_1 City_ Medina quatit 10 Gity_Renton room_bed 4 City_Maple Valley City_Sammamish sight 3 ity_lssaquah {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm imp 0.28448 oazar 0.09463 0.06691 0.05062 0.04008 0.03731 0.02532 0.02459 0.02049 0.01527 0.01319 0.01268 0.01245 0.01226 0.01089 0.01047 0.00945 0.00942 0.00908 0.00830 0.00648 0.00856 0.00545 0.00821 0.00393 0.00388 0.00379 0.00351 0.00303 capstone projectreport, aunts3716724, 8:06 PM capstone projectreport, In [163]: from sklearn.pipeline import Pipeline In [164]: def result (model, pipe_model,x_train_set,y_train_set,X_val_set,y_val_set): Pipe_model.fit(x_train_set,y_train_set) apredicting result over test data y_train_predict= pipe_model.predict(x_train_set) y_val_predict= pipe_model.predict(x_val_set) trscore=r2_score(y_train_set,y_train_predict) ‘trRMSi viscore-r2_score(y_val,y_val_predict) VARMSE=np.Sqrt(mean_squared_error(y_ val, VAMSE=nean_squared_error(y_val,y_val_predict) vIMAE=mean_absolute_error(y_val,y_val_predict) result_df=pd.DataFrame({‘method* ANSE, “MSE_val' :VIMSE,‘MSE_v1': vIMSE, train Score’ MSE, ‘MAE_tr': trMAE}) return result_df Above function will run the model and return the r2 score,rmse,mse of the model {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm al_predict)) model], ‘val score’ score, "RMSE_tr’ p-sqrt(mean_squared_error(y_train_set,y_train_predict)) ‘trMSE=mean_squared_error(y_train_set,y_train_predict) ‘trMAE=nean_absolute_error(y_train_set,y_train_predict) ‘score, 'RMSE_val" :v1 ‘trRMSE,‘MSE_tr': te zante3116124, 8:08 PM capstone projectreport, In [165]: | #Creating empty dataframe to capture results _dff=pd.DataFrame() pipe_LR = Pipeline([(‘LR', LinearRegression())]) d.concat ([result_dff,result(‘LR' ,pipe_LR,X_train,y_train,x_val,y_v pipe_knr = Pipeline({(*KNNR", KNeighborsRegressor(n_neighbors=4,weights="dista nce'))]) result_dff=pd. concat([result_dff,result(*KNNR' pipe _knr,X train,y train,X val, y_val)]) pipe_DTR = Pipeline([(‘DTR', DecisionTreeRegressor())]) result_dff=pd.concat([result_dff,result('DTR" , pipe_DTR,X_train,y_train,X_val,y -val)]) pipe_GBR = Pipeline([("GBR’, GradientBoostingRegressor(n_estimators = 200, lea rning_rate = @.1, random_state=22))]) result_dff=pd. concat([result_d#f,result( GBR’, pipe_GBR,X_train,y_train,X_val,y -val)]) pipe BGR = Pipeline([("SGR', BaggingRegressor(n_estimators-5@, oob_score= Tru e,random_state=14))]) result_dff=pd. concat([result_dff,result( ‘BGR’, pipe_8GR,X_train,y train,X_val,y -val)]) pipe_RFR = Pipeline([(‘RFR’, RandonForestRegressor())]) dd. concat ([result_dff,result(*RFR", pipe RFR,X_train,y train,X_val,y result_atf out(265): Method 22! RMSE_val—SE.val st MMSE tr OLR avieras Tar/easoeate ‘esrGsrevt0 Tes7oe7evto o7aoii2 TenesesorE=1 178r 0 KNWR 0425008 18658645160 a7e%E7er10 A7EI57e+t0 Ose8eab — s4so.iezo71. 8.987 0 IR asarate 176errarswsr s.trss0evt0 a.214800et0 aseecze saso.ezori .9er 0 oR aven4rs s2tianseezz8 14srzsrevio.as7247eetO 0820621 108834769896. 1.179 0 BOR a76est9 12478801887 1s8886G0rI0 1EEES58e+10 866456 468E7.11894 2.108 0 ER 057473 ‘270.7882 1638Hbt0vt0 e3E86tert0 assse8o ost azzse 2922 Above sequence of steps with pipeline function will run all the models and compile the scores in result_dff dataframe. We can see that the above 2 steps are concise instead of running individual models and compiling the scores as earlier. We can clearly see gradient boosting is giving better result in comparison with other ensemble methods. Also the score of 0.82 on training set indicates no overfitting of the model {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm 2sii783116124, 8:08 PM capstone proj In [166]: | #Storing results of initial data set - off result_ds1-result_dff.copy() ject report result_ds1 out [166]: Method 2! RMSE_val © MSE_val mse MM RMSE tr © LR 0.718749 197733608615 1.897057e+10 1.897057e+10 0.730112 132958.967261 1.7671 0 KNNR 0.425008 196935.451160 3.878357e+10 3.8783670+10 0.998628 9480.192071 8.987: © DIR 0.537219 176677:375867 3.121490e+10 3.1214908+10 0.998628 9480,192071 8.987 © GBR 0.782471 121129.989228 1.487247e+10 1.467247e"10 0.820821 108334,766538 1.173 © BGR 0.769319 124738.101557 1.555959e+10 1.5559590+10 0.966468 46867.181594 2.196! © RFR 0.757473 127900.773592 1.635861e+10 1.635861e+10 0.955380 54061.682258 2.922 » FEATURE SELECTION (PCA) Now, we will explore the possibilty of features reduction using PCA In [167]: df#.shape out[167]: (18287, 91) ‘e:IC:Rsor!SAUMYA SINGHIDocumen'siPP_RE_01 - House Pelee Predicton'Salt/capstone projectreport. 2619783116124, 8:08 PM In [168]: dff.columns out[168]: Index(['price’, "yr_built', "HouseLandRatio", ‘room! ‘room_bed_4*, ‘room_bed_: "room_bed_9", ‘room_bed_: "Living_measure’, “Living_measure1s", capstone projectreport, “lot_measure’, ‘ceil_measure', ‘basenent’, lot_measureis", ‘total_area’, bed_1", ‘room_bed_2", ‘room_bed_3", 5", ‘room_bed_6', ‘room_bed_7', ‘room_bed 8", 1, ‘room_bed_11', ‘[email protected]", ‘[email protected]", ‘room_bath_1.0", ‘room_bath_1.25', ‘room_bath 1.5", "room_bath_1.75", ‘room_bath_2.@", ‘room_bath 2.25’, ‘room_bath 2.5", ‘room_bath_2.75", ‘room_bath_3.0", ‘room_bath 3.25’, ‘room_bath 3.5", ‘room_bath_3.75", ‘room_bath_4.0", ‘room_bath 4.25’, ‘room_bath 4.5", ‘room_bath_4.75", ‘room_bath_5.0", ‘room_bath 5.25’, ‘room_bath_5.75', "ceil 1.5", ‘ceil 2.0", ‘ceil 2.5", ‘ceil_3.0°, ‘ceil 3.5", ‘coast_1', ‘sight_1', ‘sight_2', ‘sight_3', ‘sight_4", ‘condition_2", ‘condition_3', ‘condition 4", ‘conditions’, ‘quality 4", ‘qualitys', ‘quality_6', ‘quality 7', ‘quality_8', ‘quality_9', ‘quality_1e', "quality_11', "City_Black Diamond’, "City_Enumclaw", "City Kenmore’, ‘City_Kent', ‘City_Kirkland’, "City_Mercer Island’, "City Medina’, a, "City_Renton’, "City_Vashon’ , "has_renovated_Yes'], dtype="object') will drop the price column as it is the target variable "quality_12', “City_Bothell', “city_Fall City’, “City _Sammamish’, "City Woodinville’, “furnished_1", “City_Carnation', “City_Federal Way’, ‘City_Bellevue', *city_Duvall', “city_Issaqua "City_Maple Valley’, "City_North Bend’, ‘City Redmon “City_Seattle’, ‘City Snoqualmie’, ‘has_basement_Yes', In [169]: df_pea = dff.drop(['price'], axis = 1) In [170]: numerical_cols = df_pca.copy() numerical_cols. shape out[17e}: (18287, 90) {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm amie3116124, 8:08 PM capstone projectreport, In [171]: | # Let's first transform the entire X (independent variable data) to zscores. # We will create the PCA dimensions on this distribution. from scipy.stats import zscore # As PCA for Independent columns of Numerical types, Let's pass numerical_cols (16 numerical features) numerical_cols = numerical_cols.apply(zscore) cov_matrix = np.cov(nunerical_cols.T) print( ‘Covariance Matrix \nXs’, cov_matrix) Covariance Matrix Xs [[ 1.@0005469 @.20028185 @.84597846 ... .05257785] [ @.20028185 1.90005469 0.1663024 -0,00617414] [ @.84597846 @.1663024 1. 89005469 0.01739462] [ @.01415428 .08035946 0.01649371 -0.01445085] [ @.20094885 -0.02988448 -2.27730605 @.04524435] 0.08035946 -0.02988448, @.01649371 -0.27730605 1,00005469 -0.0056238 + -®.0056238 1.00005469 [ @.05257785 -0.00617414 @.01739462 ... [email protected] 0.64524435 1,0005469]] As we can see, near the value to 1, more the features related. {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm 0.01415428 0. 20094885 zane_anai24, 8:06 Pw capstone projectreport, In [172]: eigenvalues, eigenvectors = np.linalg.eig(cov_matrix) print(‘Eigen Vectors \nXs", eigenvectors) print('\n Eigen Values \nXs', eigenvalues) Eigen Vectors %s [[ 3.38140157e-01 -5.91272225e-02 2.10933458e-01 ... -5.27282174e-03 5.54142192e-@3 -1.67124034e-04] [ 7.12659835e-02 -4.34121260e-01 -8.88436080e-02 ... -1.68818774e-02 4,57078107e-03 -8.52995967e-03] [ 3.49772357e-01 -8.00383876e-03 -4.05156781e-02 ... -4.68496505e-03 -1,39683527e-02 2.43243338e-03] [ 1.29688720e-02 -3.8039856e-02 -3.41813657e-02 ... 1.07174062e-01 8.41737918e-02 -1.95413367e-01) [-2.50075399e-02 -3.74622282e-02 4.39476539e-@1 ... 3.99982110e-03 4,08181763e-02 2.32617269e-02] [-3.18537004e-03 -6.23000043e-04 1.02661626e-01 ... 2.84579669e-02 -1,47963772e-02 -6.62692406e-02]] Eigen Values %s [ 6.40030103e+00 4.23053272e+00 3.02200570e+00 2.36069955e+00 1.72278028e+00 1.70533047e+00 5.17634008e-02 7.8486425Se-02 1,23323929e-01 1,58239483e100 1.94704947e-01 2.10588552e-@1 2.45372409e-01 3,37764061e-01 3.52383334e-01 2.24756725e-03 9,93351422e-04 1,28503648e-04 8.54683326e-05 1.51669793e+00 3.97816689e-01 1,48400510e+00 4.25049450e-01 -5.20329656e-16 -1.83229560e-15 3.57406920e-15 1.39212554e+0@ 1.33812387e+00 5.71411667e-@1 6.48215227e-01 6.60453404e-@1 1.27455883¢100 6.90208644e-01 7.3090@85Se-01 1.22358633e+00 1.21781188¢+00 7.54916613e-@1 7.61951753e-01 7.89272221e-@1 1.19439921¢+00 1,18354682e+00 8.08765828e-01 8.31761100e-@1 1.17521503e+00 1.16073113e+0@ 8.62975337e-01 1.14847039e+00 8.79158894e-01 1,11948938e+00 1.10960276e+00 8.90644524e-@1 8.88567656e-01 9.01761603e-01 1.10493861e+00 9.16012433e-@1 9.31041146e-01 1.09143428e+00 1.08460485e+00 1.08273453e+00 1.07118893e+00 9.33793856e-01 1.06368893e+00 9.41694315e-@1 9.44273389e-01 9.49801385e-01 9.52927340e-01 1.05455299e+00 1.04955645e+00 1.04815072e+00 1.04163633e+00 9.69808694e-01 1.03813696e+00 1.03345195e+00 1.02768165e+00 1.02381893e+00 9.82887562e-01 9.81254198e-01 9.86986081e-01 1.01521697e+00 9.89390243e-01 9.93680575e-01 9.93261992e-01 1.00195909e+00 1.01363137e+00 1.01189827e+00 1.00051625e+00 1.00419597e+00 1.00622516e+00 1 -00552678e+00 1.00928050e100) {ie:1IC:1Users/SAUMYA SINGH/DocumenisiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Mtn zane3716724, 8:06 PM In [173]: capstone projectreport, # Let's Sort eigenvalues in descending order # Make a set of (eigenvalue, eigenvector) pairs eig_pairs = [(eigenvalues[index], eigenvectors[:,index]) for index in range(le n(eigenvalues))] # Sort the (eigenvalue, eigenvector) pairs from highest to Lowest with respect to eigenvalue eig_pairs.sort() eig_pairs.reverse() print (eig pairs) # Extract the descending ordered eigenvalues and eigenvectors eigvalues_sorted = [eig_pairs[index][0] for index in range(len(eigenvalues))] eigvectors_sorted = [eig_pairs[index][1] for index in range(len(eigenvalues))] # Let's confirm our sorting worked, print out eigenvalues print( ‘Eigenvalues in descending order: \nks' %eigvalues_sorted) {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm 30/1783116124, 8:08 PM capstone projectreport, [(6.400301029851477, array([ 3.38140157e-01, 7.12659835e-@2, 3.49772357e-0 1, 1.21826278e-02, 2,45688226e-01, 3.13509594e-01, 5.85711636e-02, 1.32118243e-01, 1.24188241e-01, -5.39557358e-02, -1.55322767e-01, ~7.38082454e-02, 2,69280354e-01, 6.71341866e-02, 1.76807150c-02, 8.40382554e-03, 2,25638942e-03, 3.61933268e-03, 1.42331836e-03, 5.172604760-04, ~5.638039840-03, -2.85596080e-02, -2.20914937e-01, -2.56073225e-03, ~6,59873013e-02, -6.94849495e-02, -5.10691490e-02, 2.99009064e-02, 1,96828062e-01, 7.27758185e-02, 4.47037234e-02, 6.43544191e-02, 8.98255374e-02, 3.57441062e-02, 2.39042699e-02, 1.82629160e-02, 2,35903040e-02, 9.52118613e-03, 6.36445584e-03, 6.71514053e-03, 4.22319789e-@3, -8.25472478e-02, 2.65712150e-01, 2.24314600e-02, 1.4ag@Ba64e-03, -1.59380115e-4, 1.36351362e-02, 1.21584859e-02, 2.79651538e-02, 3.46589095e-02, 2.87868177e-02, -3.59443998e-02, 1.23735329e-01, -9.23789546e-02, -5.48045043e-02, -2.04647878e-02, 5.44353495e-02, -1.55877412e-@1, -1.66909198e-01, 9.78807036e-02, 2,09856238e-01, 1.36794862e-01, 5.65981168e-02, 8.75831765e-03, 2.64286457e-01, 4.83149865e-02, -8.63848741e-03, 2.05515336e-02, 1,04425520e-02, 2.46487074e-02, -1.36322635e-02, -8.34913899e-03, 5.58571616e-03, 5.97895799e-02, 1.67219076e-02, 1.73286710e-02, 4.02906127e-02, 1.35579146e-02, 3.20240017e-03, 5.54293286e-02, 2. 79650623e-02, 1.80305747e-01, 5.49209909e-02, -5.88852015e-03, -2.5075399e-02, -3.18537004e-03])), (4.230532715020! 0.43412126, [email protected], -0.09492954, 0.13056396, -0.09447739, [email protected] , -0.42105132, @.38853903, 0.00818031, @.09903886, -2.00322975, -2.05039931, -0.0397953 , -0.01135684, -0.00195537, @.00187349, 9.00561416, -0.00403777, @.00329149, @,00403127, -0.01266106, -0.02264596, 0.00982359, 0.00641089, -0.1047631 , -.01951@81, -0.01220764, 0.1019734 , -0.03¢86617, 0,0212307 , 0.04710655, 0.04081646, -0.00117921, 0.00207246, @.00066914, -0.00061282, -2.00468101, -0.00095167, .00073386, 0.000858 , -0.00237293, 0.15398148, 0.03143851, .20703993, @.0235899 , -0.04163645, -2.02784399, -0.01428115, -2.01787219, -0,04086905, -0.02070046, 0.21897162, -8.20078997, -0.05272298, -0.00838391, -0.0126264 , -0.02109097, -8.07477947, 0.10406576, @.01781302, -0.03510238, -2.02017662, -0.00081656, -2.00543061, -0.12229314, -0.02405586, -0.01865954, -0.03518784, -0.00616237, -0.02599962, -0.02240527, -0.06022892, 0.0647382 , -0.03107846, -0.03563074, -0.05546802, 0.03231613, -0.02937766, -0.07517532, -0.03530176, 0.00217373, -0.04298141, -0.03445572, 0.1795959 , 0.03400426, -0.04414761, -2.03803986, -0.03746223, -2.000623 ])), (3.0220056962108997, array({ 0.21093345, -0.08884361, -0.04051568, 0.459233 1, -0.20948573, 0.06094638, -0.08769888, -0.04338803, 0.18999279, -0.03429171, -0.04303535, -0.14081992, @.09630094, 0.14413304, 9.08075213, @.03087204, 0.021635 , 0.02874812, @.01352604, .01099741, -0.00046969, -0.0253926 , -0.12810462, 0.00801375, -0.00546132, @.08028745, 0.05174082, 0.02460706, -0.17141934, .10063382, @.10513547, 0.08968952, 0.09751819, 0.04988048, 0.04180795, @.0360652 , @.04631527, 0.01510422, 0.00364895, 0.00749181, @.00784258, 0.14066273, -0.13476605, 0.058906 , 0.01324093, @,00143509, 0.04433479, 0.0745123 , @.11139792, @.10282392, @,08280422, -0.02017733, -0.15120561, 0.08089571, 0.1413166 , -0.03014322, [email protected], -0.07345488, -0.00271769, 0.02248996, @.03418131, 0.03990794, 0.02663322, 9.01611412, 0.05546425, 0.03544567, -0.03015266, -2.01667398, -0.0401713 , -2.05250393, -26648246e-02, -06324757e-02, -07319938e-01, -29688720e-02, ‘1, array([[email protected] wnpuneee 2722, {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm sure3116124, 8:08 PM -0,04129285, -0,10458144, -0.056401 , -0.05563353, (2.3606995503606907, 8, -0.20077718, 0.05062147, 0.26402065, 0.01216159, 0.01828875, -0.17577786, 001273173, 0.03028663, 0.20706833, -0,00721205, @.11456154, @.07277432, 224099786, -0,03052305, -0,00267958, -0,10262501, -0,03335646, @.00513287, (1.7227802802980245, 2, 5.16876256e-02, 4,84901490e- 2.21522986e~ -6.46767863e- 2.57882460e- 1.16979339e~ 1,00421405e~ 1,05193372e~ -2,18796507¢- 5.98256022e~ 4,41525866e- -5,32191501e- 1.73155089e~ -5.78750293e- 5..27653264e~ 0.03835547, 09678647, 1074207 -0.01259858, -0.02191463, 010503604, -0.01547844, .0258032 , -0.00839736, -0.00625977, 304967, 0 -0 -0.02118915, -0.01955768, -0.05468041,, -0.00285831, array ([-6.00720377, 0.01439989, -0.27661743, -0.00804137, 0.09307033, -0.06178217, 0.05388027, 002775791, @.10759155, @.10292123, @.0672181 , @.13346116, 216208952, 001273904, @.02118784, -0.05898896, -.03231935, 003516863, array([-2.75891943e-02, 02, 04, 03, 22, 1, 22, 21, 21, 22, 22, 22, 22, 03, 02, -0,049328 @.07094793, 044510897, -.00250479, -0.00796351, e.e0asa758, 012430449, .044399 -0.10453813, 025531019, 1.72202394¢-02, 1.75531777e-01, ~6.99276559e-02, -1.38633482e-02, 1.31501431e-02, 2,03529963e-01, -7.22222733e-02, -9.70041136e-03, 1,12053923e-02, -1.57737740e-02, -1,82694555e-02, -3.03469013e-03, 6.17820745e-02, -5.39089932e-03, 2.06938171e-02, -1.19344878e-01, 1,86698328e-02, 3.70161361¢-02, 8.36701642e-02, -5.79924748e-02, 8.151025610-02, -2.36634142e-01, -3.34803092¢-02, 2.773702460-02, 3.65080662e-02, 3.01403507e-02, -3.87528058e-02, 2.91986845¢-02, -2.82725675e-02, 8.14799322e-02])), (1.7053304684180373, array([-9. capstone projectreport, -0.04979088, -0.04044887, -0 -.10198589, @.00537009, @ -0.08192034, -0.05254563, @ -.03418137, 0.43947654, @ .00224618, @.00073809, -0.07253922, @ @.05624118, 6.02404274, 0 @.00257074, -0.00579735, 0 @.32371789, @.01903027, -@ -0.14789745, -8.06749212, @ 0.072458 , 0.05551253, @ @.0149579 , -0.00807084, 0 @.00949538, .0310457 , -8 2.013677 , @.03206102, @ @.02340023, -0.05185769, @ @.2975113 , -8.2129159 , -8 @.06896703, 0.00896484, 0 -0.03055353, 0.01560851, -0 [email protected], -0.02238202, -0. [email protected], 0.01533014, 0 [email protected], 0.10004965, @ -0.02673597, -0.16413819, 0 6.22663979e-02, 1.03722364e-01, 5. 7.48912126e-02, 3.87175860e-02, -1.18895907e-02, -2.22415917e-02, -8,15912951e-02, 7,362202960-03, 2.37479559e-02, 1,58457297e-02, -1,@5065660e-02, 3.82937199e-01, 3.50285374e-01, -6,94195526e-02, -3,43044936e-01, 3.91109747¢-02, 1.42157352e-02, -7.91148709¢-03, 2.24997611¢-02, 6.e6045961e-03, -1.62743839e-02, 1.74298356e-01, 1. -2. 1. 1. “1. -8. 4. 1. 4. 3. 2. “1. 0.20682403, @.0279412 , -0.38919037, @.e1319458, -0.02695272, -0.01146529, 207892846, @.01198528, 005859151, 2.00308563, 2.06603838, 0.01134168, -2. -0.0100558 , -@ -0.0099384 , @ 0.04413847, 0. -0.12597964, -@ 0.04778133, 0. 000248636, @ 0.02176298, 0. 0.02034035, 0. {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm 0.07573216, 021319, 0366488 26984296, |.10266163])), 0.145471 14716631, 00526146, 0020183 , 05286308, 02029172, 03184281, 00573534, 05454233, 24935804, 01734833, 23715036, |. 30668776, 02733405, 04074563, 04045805, |. 16962266, -@8959676])), -5.87154230e-0 34730987e-02, -2,40660821e-02, 78451840e-02, 37987660e-03, 4.43677778e-02, 25976552e-01,, 20741003e-02, 610449220-03, -9.48287976e-04, 78509029¢-03, 2.087404160-02, 35659169e-02, 5@096333e-01,, 55724591¢-01,, 62507673¢-03, 65573237e-03, 3.45710951¢-02, -6.09415475¢-02, 2.43959995e-02, 11980582e-02, 4.41241036e-03, 03214826, 00243799, 00592628, 07172762, 10725771, 04080463, 01644248, 14312427, 07405128, 3276_anai24, 8:06 Pw @. 0. @. 0. 0. 0. 8. 0. (1,5823948341725016, 25196789, 22796998, 12769751, 03815552, 00507272, 03981033, 03969387, 01284676, 7, -8.01142927, @. -0. 0 8. @. -0 -0 0 @ @. -0 @ @. @. @. 0. 0. (1.5166979298471606, 04732042, 05834357, 00748711, 01127987, 01776535, 08995488, 00506415, 01695773, 29645225, 11947425, 07828787, 13986082, -17926009, 02650792, 03097681, 06511834, 03140204, 6, -0.0823675 , 0. 0. @. 0. -0. @. -0. @. -0. @. 0. 0. 2. @. @. 0. @. (1.4840050960170488, 01772739, 6798112, 02817224, 02279888, 03952647, 01929373, e1iss9s8, 01212473, 20819852, 39971337, 12236071, 1¢684081, 27205484, eaaoass , 1882041, 21529219, 25796855, » 0.03689281, 0 -0 -0 @. @. 0 0 0 0 00897316, 14582129, 04212032, 02341597, 2429155 02356482, 01161743, 0362343 01584929, capstone projectreport, -0.03486212, 0. -0.07348146, -@. 015962844, 0. 0.00385181, 0. 0.00141885, -0. 0.05873867, -0. -0.04650538, -0. 0.07491223, 0. array([-0.02579091, -0.10323643, -0.07678036, -0. @.22674486, -2. -0.06773762, -@.. -0.07419635, -@. -0.01940218, 0. @.07290427, -@. -0.04276741, 2. @.06618993, -2. 0.1060034 , -0. -0.01927545, -@.. .00855355, 2. @.02992472, @. 0.00538621, -2. -0.0351676 , @. -0.03501136, -@. e.cea7saas, 2. 0.02211096, -@. array([ 03512953, -¢.05998021, -0.04739566, -0. -0.1294a801, 0. 0.02324206, 0. -0.08033275, -0. 0.13790474, -0. -0.1169406 , -0. 0.08773861, -0.. 0.23575178, 0. 041200847, -0. -0.05494921, -0. -0.02983097, -0. -0.04685742, 2. -0.02058402, -0. -0.03878011, -@. -0.01572402, @. -0.05176724, @. @.11527288, -@. -0.04864121, -@. -0.01310922, 2. -0.00445327, -@. 0.51007806, -@. 0.02903634, -@.. @.00522301, -@.. -0.06236782, -0. -0.02652322, 0. 0.00656723, 0. 21168428, 29426926, 10054937, 01635949, 0511901, e4sa2as2, 12865868, 04722783, 10237832, 12169029, es1sa6s4, 03121154, 10655432, e4as1162, 1211364, 1794634, 02057187, 48625332, 24908957, 01678794, 05599342, ee16s846, 0624873 , 2232029 , 04893219, 4986913, 22644149, 03927625, 08179817, 1450146 , 12399375, 1351911, egasasaa,, 10110496, 07489579, 15292934, 03030548, 01707269, 8398554, 13624518, 05367334, e4911701, 27710089, 06974777, 02754135, 16105521, 02010313, 00311867, 2080576 , 03201913, 1599246, -0.15449296, .18688518, 0.0230538 , 0.02aaaze2, -0.0331081 , 0.01465507, 0.08653509, 0.05148136, @.11412126, [email protected],, 0.0104614 , 0.04688368, 9.007984 , -0.00111173, -0.06485362, 0.07672332, -0.e020006 , 2.a84a0686, -0.20662059, -0.00525954,, -0.0716547 , 8.09514562, -0.00370757,, -0.00802633, -0.14940765, 0.017378, -0.07273589, .01264389, 0.02656226, 0.129426 , 0.02649847, 0.0249923 , -0.02308546, -0.11640363, -0.00541504, 0.2874068 , -0.03453339, -0.05880667, -0.07418455, -0.02676439, -0.00107446, -0.15494206, 0.05563915, -0.01969408, .01184245, 0.03972466, 0.05042578, -0.02191703, -0.02258844,, -0.00857236, -0.04447129,, {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm -0 -6 @ -6 -6 @ a @ 0.06233093, -8. -8. -8. e. -8. 03052541, 03264406, 18427271, 02221682, 0765737 08991486, -14152616, 08164325, 07842145, -11001634, 05344581, -12861492])), -0.1215128 -0 -0 @ @ @ @ @ -0 -0 @ -0 -0 0. 10666813, -0. a. a. @. a. a. e. “0. -0. e. -0. -0. e. -2. -8. -0. @. array([-0.01647521, [email protected], [email protected], @ -0 -0 @ @ -6 -0 -6 -6 11058998, 28658732, 21152038, 0144477 01035132, 01155537, 08155747, |.@8652505])), 0. 1563163 06317361, 29161384, 3208893, 4296296, ese09aa4, 07612813, 26650317, 0187637 , 07572413, 0847142, 03841585, 21544068, 8859317, 04605829, 15577549, 11035494, 11326956, 1668012, 2799173, 5427701, 0264562, 11725425])), |. 38743286, 06385795, 0039486 , 0224396 00409658, 04233229, 02122084, 02097777, 00803429, 0.0359174 sare_anai24, 8:06 Pw capstone project -0.01061251, @.10817242, [email protected], 0.08557665, 9.0460519 , 0.47181799, @.16525984, -2.20280355, 0.17355343, [email protected], 0.1237945 , -0.04323163, -2.03597109, -0.00419303, 0.08125443, -0.00628357, @.04920573, [email protected], 0.17418589, -8.21709192, e. 8. -2. port @.00150704, 0.01570768, -0.04528673, 0.04077768, -0.0136485 , @.07800799, 0.04705179, .04622736, -0.02957357, [email protected], -0.04611269, -0.03345367, 0.0315392 , -0.03166596, -0.06843772, @.01994364, 0.02590641, -2.00907726, 0.03793356, -2.06581374])), (1.,3921255385114626, array([-7.44798782e-02, -2.04057215e-02, -1.70213024e-0 1, 1,59969471e-01, 1,16745291e-01, | 1.22570505e-02, -8.60491578e-03, -3,35444803e-02, -1,63747860e-02, -8.81691725e-02, 2.29829604e-01, -1.91980144e-01, 2,00935995e-01, -9.24163310e-02, -3.58916386e-01, -1.55938407e-01, ~8.01419058e-02, -1.67543214e-@1, -1.585900800-02, -3.33998688e-02, 1,420906600-02, -1.01937834e-@1, 3.97490694e-02, 3.114868010-02, =3.28015525e-02, 1.03256895e-01, -1.58237285e-01, 7.35541952e-02, 8.49211152e-02, 3.08494796e-02, -1.74544509¢-01, -8.39287128e-03, 2.78487996e-02, -9.45944424e-02, -1.317263000-01, -4.56532786e-02, -2,77369029e-01, -1.65078423e-02, -1.47011482e-01, -1.20544364e-01, ~1,24713634e-01, -2.12675713e-01, -1.59140924e-06, -1.70492395e-01, ~4.71749996e-02, -2.64371816e-02, 1.93074103e-01, 5.10871246e-03, ~4,51392455e-02, -1.12503289e-02, 1.79238763e-01, -3.64441742e-02, 4,17966706e-02, 9.30520985e-03, -7.65341683e-02, -1.14372474e-01, ~3.05317310e-02, 3.57304046e-03, 6.67812385e-02, -1.18929855e-01, 1.15e80118e-01, -2.72141¢11e-02, -6.35371089e-02, 1.76916211e-03, 7,65631431e-02, 1.08834391e-02, -3.97408678e-02, 9.80381325e-03, ~7.29636808e-02, -3.18072086e-02, -1.22301352e-01, -4.27613032e-02, 5.05198151e-02, 2.49743560e-02, 5.62910547e-02, 8.95831493e-03, 3.61783200e-02, 3.30025629e-02, -7.22496957e-02, -5.24225212e-02, ~3.43834525e-02, 3.517116@1e-02, -1.93835937e-02, 3.89598662e-02, ~3.88440865e-02, -4.38094026e-03, 7.82669060e-02, 3.50383193e-02, 1,99411248e-01, -1.66581180e-01])), (1.338123871891159, array([-0.005 48775, -0.03843159, -0.03061153, 0.04333244, 0.13993639, -.03706519, -0.02074415, [email protected], 6.03333833, [email protected], @.22987535, -0.21522097, -0.06384675, 0.08397588, 0.34627929, =.04155743, 0.15962213, 0.17622303, 0.03179456, -0.00631348, @.01499474, -0.03142747, 0.07360389, 0.03405806, 0.05415617, -0.04012176, -0.21879165, @.10111445, -0.06312339, -0.02052279, @.07695396, -0.03795474, 0.02675084, 0.09266905, .02364962, @,00294061, 0.32037628, 0.04085849, 0.22018367, 0.15760833, -0.87506903, -0.26341749, 0.045752 , 0.046454 , 0.0182172 , -0.00876758, @.02950111, -8.07382824, -0.09627891, -0.03991851, @.05040326, 0.00198497, -2.07373555, 0.24160928, -2.26190678, -0.01717238, 0.05946142, -0.07260429, 0.15436931, -0.14236139, -0.00075114, -0.02015509, 0.12668078, -0.02670258, 0.01202094, 0.07757646, -0.01062797, -2.00859107, -0.03864637, -2.04924234, -0.12007782, -0.04770347, 0.01634402, 0.02039453, 0.00370991, 0.067483, 0.07424346, 0.01216043, 0.06891506, -2.01504565, -0.05261312, 0.05696704, 0.06306901, -0.0986121 , -0.09694006, -0.02580625, 0.00048615, 0.025849 , 0.04712735, -0.18562838])), (1..2745588269563284, array([-1.35554753e-03, -6.07757998e-03, 1.94705449e-0 2, -3.64873032e-02, 2.57177745e-02, -1.73766948e-03, -7.66767391e-03, -5.99583989e-03, 1.28081101e-02, 3.13125071e-03, 7.72361809e-02, -9.08219621e-02, ~3.02525331e-02, 1.04350856e-01, -1.07304336e-01, 5.79268206e-01, ~1,92585215e-02, -1.43456269e-01, 1.44049771e-02, 5.54833874e-03, ~3.02947376e-03, 1.53579737e-02, -1.61490155e-05, 1.64651561e-03, {ie:1IC:Users/SAUMYA SINGHiDocumentsiPP_RE_01 - House Price Predillon/Sol t/capstone projec report. Nm 3an78