AI/ML Detail Roadmap

By: ANANDA RIMAL

Introduction
 1. Git and
Github

Link
2.Linkdein
 3 Python
Step1:
•✅ Variables, Data Types (int, float, str, bool)
•✅ Operators (arithmetic, logical, comparison)
•✅ Conditional Statements (if, elif, else)
•✅ Loops (for, while, break, continue)
•✅ Functions (def, return, *args, **kwargs)
•✅ String manipulation
•✅ Input/Output operations
•✅ Exception Handling (try, except, finally)

🛠 Practice: Build small CLI tools (calculator, quiz app)

🟡 Step 2: Data Structures & Built-ins
📌 Goal: Efficient data handling
✅ Lists – Ordered, mutable, indexable, supports slicing
✅ Dictionaries – Key-value pairs, fast lookup, dynamic
✅ Sorting – sorted(), .sort(), custom key support
✅ Slicing – list[start:end], list[-n:], flexible access
✅ Searching – in for lists, .get() for dictionaries

🛠 Practice: Use user input to create a list and dictionary,

then apply sorting, slicing, and search operations
🔵 Step 3: Object-Oriented Programming (OOP)
📌 Goal: Organize and structure code like a pro.
•✅ Classes and Objects
•✅ Constructors (__init__)
•✅ Instance vs Class Variables
•✅ Inheritance, Polymorphism
•✅ Encapsulation, Abstraction
•✅ Magic Methods (__str__, __repr__, etc.)
🟡 Step 4: Functional Tools & File Handling
📌 Goal: Write clean, efficient, and reusable code

✅ lambda – Anonymous, one-line functions

📌 Used with map(), filter(), sorted()
✅ generators – Memory-efficient iterators using yield
📌 Useful for large data processing
✅ iterators – Objects with __iter__() and __next__()
📌 Used in loops and generator expressions
✅ file handling – Read/write using open(), with
📌 Modes: 'r', 'w', 'a', 'rb', 'wb'
 Step5:

Link for python and flask

Estimated :2 months
 4 MATHEMATICS FOR ML
Linear Algebra
Vectors – What to Study for AI/ML
•What is a vector
•Vector representation
•Vector addition
•Scalar multiplication
•Vector notation
•Dot product
•Cosine similarity
•Cross product
•Euclidean distance
•How a feature set represents a vector
•Concept of hyperplane
 Matrix

Matrix – What to Study for AI/ML

•represent a matrix
Link1: for
•Types of matrices (square, diagonal, linear algebra
identity, symmetric, orthogonal, sparse)
•Transpose of a matrix
•Determinant of a matrix Link2:for linear
algebra
•Inverse of a matrix
•Rank of a matrix
•Linear independence of rows and
columns
•Null space and column space
•System of linear equations using matrix
form (Ax = b)
•Gaussian elimination method •Eigenvalues and
•Row echelon form and reduced row eigenvectors
echelon form •Diagonalization of a matrix
•Matrix multiplication •Singular Value
•Scalar multiplication Decomposition (SVD)
•Matrix addition and subtraction •Matrix as a linear
•Matrix factorization methods (LU, QR) transformation
•Matrix representation of
dataset (features and
samples)
•Matrix operations using
NumPy
 Calculus

•Local minima and

maxima
Calculus – What to Study for AI/ML
•Convex and non-convex
•Limit
functions
•Derivative
•Learning rate
•Slope
•Cost/loss function
•Gradient
•Jacobian matrix
•Gradient descent
•Hessian matrix
•Chain rule
•Multivariable calculus
•Partial derivatives
•Function
•Optimization
approximation
•Taylor series expansion
•Backpropagation
 Probability and
statistics for ml
🔹 1. Basics of Statistics
🧠 Understand what statistics is and why it's used in ML
•Definition of Statistics
•Types of Statistics
• Descriptive Statistics
• Inferential Statistics
•Types of Data
• Qualitative (Categorical) vs. Quantitative
(Numerical)
• Discrete vs. Continuous
• Scales of Measurement: Nominal, Ordinal,
Interval, Ratio
2. Descriptive Statistics
3. Probability
📊 To summarize and visualize data
🎲 Foundation of probabilistic models and inference
•Measures of Central Tendency
•Basic Concepts
• Mean, Median, Mode
• Sample space, Events
•Measures of Dispersion
• Independent & Dependent Events
• Range, Variance, Standard Deviation,
• Conditional Probability
Interquartile Range (IQR)
• Bayes’ Theorem
•Shape of Distributions
•Rules of Probability
• Skewness (left/right)
• Addition Rule
• Kurtosis (peaked/flat)
• Multiplication Rule
•Data Visualization
•Combinatorics (briefly)
• Histograms
• Permutations
• Box plots
• Combinations
• Scatter plots
• Bar charts
• Pie charts
•Z-score & Standardization

Estimated 2 months for

mathematics
🔹 4. Random Variables 🔹 5. Mathematical
📌 Key for understanding distributions and
expectations Expectation
•Definition 📈 Core of understanding
•Types expected behavior
• Discrete Random Variable
• Continuous Random Variable •Expected Value (Mean)
•Probability Mass Function (PMF) – for •Variance and Standard
discrete Deviation
•Probability Density Function (PDF) – for
continuous •Linearity of Expectation
•Cumulative Distribution Function (CDF) •Moments (optional unless
needed)
 7. Inferential Statistics
6. Probability Distributions 📌 Drawing conclusions from data
🧠 Most-used distributions in ML •Population vs. Sample
🔹 Discrete Distributions: •Sampling Techniques
•Bernoulli Distribution • Random, Stratified, Systematic, etc.
•Binomial Distribution •Central Limit Theorem (CLT)
•Poisson Distribution •Estimation
•Geometric Distribution • Point Estimation
🔹 Continuous Distributions: • Confidence Intervals
•Uniform Distribution •Hypothesis Testing
•Normal (Gaussian) Distribution – VERY IMPORTANT • Null vs. Alternative Hypotheses
•Exponential Distribution • p-value
•Log-normal Distribution • Significance Level (α)
•Beta & Gamma (optional, but helpful in Bayesian • Type I and Type II Errors
ML) • t-tests (1-sample, 2-sample), z-tests
• Chi-Square Test
• ANOVA (Analysis of Variance)
 8. Correlation and Regression
📉 Understanding relationships between variables
•Covariance vs. Correlation
•Pearson and Spearman Correlation Coefficients
•Simple Linear Regression
•Multiple Linear Regression
•Assumptions of Linear Regression
•Multicollinearity (Variance Inflation Factor)

Stats notes repository on

github
For statistsics tutorial watch
campus x
 5.Data
Analysis

Link:
click on
this
 Pandas

Link for pandas

Used for data

analysis
Matplotlib
Seaborn
Plotly

Used for data

visualization

Again check for

campus x dsmp free
course available on
youtube
Data
Gathering

Web
Scrapping
Link:follows campus x
Eda dsmp course for
eda,data gathering and
web scrapping
Sql
Link1:click
on this

Link2:click on
this
Now Python and Data Analysis part is
complete

Estimated 1.5 months for data analysis

6.Machine Learning

Data Handling
•Data preprocessing
•Handling missing values
•Encoding categorical data
•Feature scaling (normalization,
standardization)
•Train-test split
•Pca
Supervised Learning
 Supervised Learning
Algorithm Types

1.Regression 2.Classification
Linear regression Logistic regression
Polynomial regression Support vector machine
Naïve Bayes
Decision tree
Gradient Boosting
Lasso regression Random forest
Ridge regression K nearest neighbors
Elastic regression
Unsupervised Learning
Algorithm
Unsupervised Learning Algorithm
•Basic concept of unsupervised learning
•Types: Clustering, Dimensionality Reduction, Association
Rule Learning
•K-Means Clustering
•Hierarchical Clustering: Agglomerative, Divisive
•Density-based Clustering: DBSCAN
•Principal Component Analysis (PCA)
•Linear Discriminant Analysis (LDA
Reinforcement Learning
 Reinforcement Learning
– What to Study
•Basic concept of
reinforcement learning
•Agent, environment,
state, action, reward
Estimated time 3 months
 Scikit Learn

Scikit-learn
📌 What is Scikit-learn?
•A powerful Python
library for Machine
Learning
•Built on top of NumPy,
SciPy, and matplotlib
•Provides tools for data
Link for ml course by
preprocessing, model
campus x building, and evaluation
Link 2 :ml by
Andrew ng
7.Deep Learning
Deep Learning – What to Learn
(Beginner Level) CNN (Convolutional Neural Network)
•Deep Learning is 95% about Neural •How images are classified using CNN
Networks •Convolution, Pooling, Flattening, Fully
•Learn how Neural Networks work Connected Layers
•Use cases: image classification, object
•Understand Forward Propagation
detection
•Understand Backward Propagation
(Backpropagation)
•What is Gradient Descent and how
weights are updated
•Activation Functions: ReLU,
Sigmoid, Tanh
•Loss Functions: MSE, Cross-Entropy
•Epochs, Batches, Learning Rate
 Other Topics for
Beginners
RNN (Recurrent Neural •Overfitting and
Network) Regularization (Dropout,
•Designed for sequence L2)
data like text and time- •Optimizers: SGD, Adam
series •Model Evaluation:
•Concepts of memory Accuracy, Precision, Recall
and feedback in •Hands-on with
networks TensorFlow or Keras
•Learn LSTM and GRU •Build your first Neural
(advanced types of Network from scratch
RNNs) •Practice with image and
text datasets (e.g. MNIST,
IMDB)
 Link 1: deep learning by alexander amini

Link2:deep learning by Andrew ng

Estimated time 3 months

8.Natural Language
Processing
 🧠 NLP for Beginners – Key Topics
1.Text preprocessing
Coursera course 2.Tokenization
on nlp click here 3.Stemming and Lemmatization
4.Bag of Words & TF-IDF
5.Word Embeddings (Word2Vec, GloVe)
6.POS Tagging & NER
7.Text Classification & Sentiment Analysis
8.Transformers (BERT, GPT)
9.Libraries: NLTK, spaCy, Hugging Face
10.Practice on datasets (IMDb, Twitter)
9 Computer Vison
🧠 Computer Vision for Beginners – Key Topics
1.Image basics (pixels, RGB, grayscale)
2.Image preprocessing (resizing, normalization, Computer
augmentation) Vision by
Murtaza
3.Image classification
workshop
4.Convolutional Neural Networks (CNNs)
5.Object detection (YOLO, SSD, Faster R-CNN)
6.Image segmentation (semantic & instance)
7.Transfer learning (using pre-trained models like Computer
VGG, ResNet) vison by
8.Face detection & recognition computer
9.Libraries: OpenCV, TensorFlow, PyTorch, Keras vision
10.Practice with datasets (MNIST, CIFAR-10, COCO) engineer

Course on
coursera
🧠 10 Generative AI – What to Study
1.Introduction to GenAI & use cases
2.Language Models (GPT, LLaMA, Claude,
etc.)
3.Text Generation (prompting, fine-
tuning)
4.Diffusion Models (for image generation)
5.Generative Adversarial Networks (GANs)
6.Prompt Engineering techniques
7.Retrieval-Augmented Generation (RAG)
8.Evaluation metrics (BLEU, ROUGE, FID)
9.Tools & Libraries: OpenAI API, Hugging Gen ai course

Face, LangChain
 link
Kaggle

For different
kind of
datasets
Thank you