🧠 6–8 Month Machine Learning Learning Plan (Merged from 4 University-Level Free Courses)
Courses Merged:
• Andrew Ng's Machine Learning (Coursera)
• NPTEL IIT Madras: Introduction to Machine Learning
• Columbia University: Machine Learning (edX)
• Stanford CS229: Machine Learning (YouTube)
Objective:
• Build core ML foundation (theory + coding)
• Integrate university-level math rigor
• Apply ML on real datasets (Kaggle)
• Develop portfolio-ready projects
📅 Month 1 (Foundations)
Focus: Supervised learning, regression, Python + math intuition
• Week 1–2:
• Andrew Ng Coursera: Week 1–2 (Linear Regression)
• NPTEL IIT: Lectures 1–3
• Project: Predict student scores using Linear Regression
• Week 3–4:
• Andrew Ng: Week 3 (Logistic Regression)
• NPTEL: Lectures 4–6
• Coding: Implement Linear + Logistic Regression in NumPy
• Kaggle: Titanic Survival Prediction
📅 Month 2 (Core ML Algorithms)
Focus: Classification, SVMs, Decision Trees, Bias/Variance, Evaluation
• Week 5–6:
• Andrew Ng: Week 4 (Regularization), Week 5 (Neural Networks intro)
1
• NPTEL: Lectures 7–10 (SVM, Trees)
• Project: SMS Spam Detection (Naive Bayes / Logistic)
• Week 7–8:
• Columbia edX: Probability Review + Naive Bayes
• Stanford CS229: Lecture 2–3 (Logistic + Generative Models)
• Coding: Implement Decision Tree using scikit-learn
• Kaggle: Heart Disease Prediction
📅 Month 3 (Unsupervised Learning + Deepening Theory)
Focus: K-Means, PCA, GMM, feature engineering, dimensionality reduction
• Week 9–10:
• Andrew Ng: Week 8 (Clustering, PCA)
• NPTEL: Lectures 11–13 (K-means, PCA)
• Columbia edX: Gaussian Mixture Models
• Project: Customer Segmentation (K-Means)
• Week 11–12:
• Stanford CS229: Lectures 4–5 (Kernels + SVM math)
• Coding: Implement PCA and K-Means from scratch
• Kaggle: Mall Customers Dataset
📅 Month 4 (Neural Nets + Evaluation + Projects)
Focus: Backpropagation, tuning, evaluation, confidence, ROC, project structuring
• Week 13–14:
• Andrew Ng: Week 6–7 (Neural Nets + Backprop)
• Stanford CS229: Lecture 6 (NNs deeper dive)
• Coding: Build 2-layer NN using NumPy
• Project: Digit Recognizer (MNIST - Kaggle)
2
• Week 15–16:
• Columbia: Evaluation Metrics, ROC, AUC
• NPTEL: Lectures 14–16
• Project: Image classifier (CIFAR10 with PyTorch)
📅 Month 5–6 (Advanced Topics + Capstone)
Focus: PGM, ensemble methods, deployment, review, real-world pipeline
• Week 17–18:
• Stanford CS229: Lecture 7–8 (PGMs, Ensembles)
• NPTEL: Dimensionality Reduction / Ensemble Review
• Coding: Implement Random Forest + AdaBoost (scikit-learn)
• Week 19–20:
• Final Capstone Project:
◦ Choose real-world dataset (finance, health, NLP, etc.)
◦ Complete pipeline: cleaning, training, tuning, evaluation, deployment (FastAPI or Streamlit)
✅ Weekly Routine (Recommended)
• 3 hrs → Watch lectures (1–2 sources max)
• 2 hrs → Coding from scratch or with scikit-learn
• 2 hrs → Kaggle projects / portfolio building
• 1–2 hrs → Notes + revision
📦 Final Deliverables:
• GitHub repo with 5–6 ML projects
• Capstone end-to-end deployed project
• Structured notes from all four courses
• Resume-ready ML portfolio
can you also give me weekly projects to build , something that show i can solve a real world problem. attract
firms like citadel, spacex, drdo, mcstanley