Scribd
Upload
15 views

Assignment 1 (1)

Uploaded by

shafaq tanveer
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
15 views

Assignment 1 (1)

Uploaded by

shafaq tanveer
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

DS-323 Machine Learning for Data Science

Assignment 1
Name: ………………. Due Date:25-12-2024
Roll#: ………………. Total Marks: 50

Dataset Description:
The time domain dataset includes the instantaneous values of three-phase current
signature data of loaded induction motor with healthy and unhealthy conditions. The
motor with various faults were operated under different load conditions to observe their
impact on current signatures. The motor faults include inner-race and outer-race bearing
faults of (i) 0.7mm, (ii) 0.9mm, (iii) 1.1mm, (iv) 1.3mm, (v) 1.5m, and (vi) 1.7mm. The
bearings with different severity levels of faults were operated under the load conditions
of 100W, 200W, and 300W. In addition, the motor health condition was also experimented
under broken rotor bar (BRB) fault with 100W and 300W loads. There are a total 39
datasets including the three-phase current data acquired under different motor health
conditions. The data was acquired at the sampling rate of 10 kHz at the rate of 1000
samples per channel using non-invasive current sensors. Seven folders containing the
different induction motor datasets on different loads.
Note: This is 14 class classification problem i.e., healthy, unhealthy, unhealthy due to bearing
faults (inner and outer race) and unhealthy due to rotor broken fault on different loads.
Question#1: You have to implement K-NN algorithm on the above-mentioned dataset with
the following steps:
• Data preprocessing: In the dataset, the total number of samples in each file are more
than 100,000 and any block of 1000 samples can be used to label the corresponding
healthy and unhealthy output. Create a new csv file for training and testing your
model in which each row containing 1000 samples with assignment of the output
label healthy/unhealthy.
• Calculate the Euclidean distance
• k-NN Model Implementation without build-in libraries.
• Apply hold-out and 10 cross validation method for training the model.
• Calculate the classification measures. (Accuracy, Recall, Precision, Sensitivity,
specificity, F1 Score)
• Find the optimized K value and plot cross-validation and test accuracy based on the
K values. You can tune your model on multiple K iterations.
Question#2: Apply the data dimensionality reduction technique PCA on the dataset and
implement all the subparts of question1.
Question#3: Which model gives the best accuracy and why?

You might also like