Machine Learning Intro
Course Duration: 32 hrs
Trainer: Chandrashekar
Training Date: Nov 5th, 6th, 7th, 8th (Thursday to Sunday, 8hr each day)
Contents
| Focus Area | Topic | Day # |
|
Data Science and ML Intro |
Pretest |
1 |
| What is DS, and how it is related to ML, AI | ||
| DS tasks | ||
| Collection | ||
| Storing | ||
| Processing and Feature Extraction | ||
| Intro to ML | ||
| Latest Trends in ML and AI | ||
| Calculus Refresher | ||
| Linear Algebra Refresher | ||
| Supervised, and Unsupervised | ||
|
KNN Concepts |
K Nearest Neighbor | |
| Coding Demo in Native Python and sklearn | ||
| Scikit Learn framework KNN | ||
| Probability Refresher |
| Bayes Theorem, and Practical Problems | ||
|
KNN Handson |
Naive Bayes Classifier |
2 |
| Its use case in NLP (part 1) | ||
| Coding Demo using NB | ||
|
Decision Tree (Limitations of DT) |
Decision Tree (Prediction, Entropy method, Construction) | |
| Construction of Decision Tree for a sample problem | ||
| Coding Demo Sklearn Decision Tree | ||
| What is overfitting? Why is Decision Tree so sensitive to it? | ||
|
Linear Regression |
Linear Regression | |
| Sample problem | ||
| What do we measure and how? | ||
| Linear Equation | ||
| What do we minimize, and how do we minimize | ||
| Calculus in Linear Regression | ||
| Gradient Descent | ||
| Linear regression Algo | ||
| Problem Solving in Linear Regression | ||
| Variance, and Bias | ||
| Hyperparameters in Linear Regression | ||
| Sklearn Framework for Linear Regression | ||
| Coding Demo Coding of Linear Regression | ||
| Linear Regression Handson | Coding Problem Native Python Coding of Linear Regression: (Practical Problem) |
3 |
|
Logistic Regression |
Logistic Regression | |
| What do we measure, and how? | ||
| Equation | ||
| What do we minimize, and how do we minimize? |
| Cost Function | ||
| Gradient Descent | ||
| Logistic Regression Algo | ||
| Coding Demo using Logistic Regression | ||
| Hyperparameters in Logistic Regression | ||
| Multiclass Classification using Logistic Reg | ||
|
Neural Network |
Neural Network | |
| Concept of Neurons | ||
| Matrix Workout | ||
| How it generalizes earlier Parametric Models | ||
| Equation | ||
| Forward Propagation | ||
| Forward Propagation Workout | ||
| What do we minimize, and how do we minimize? | ||
| Cost Function | ||
| Backward Propagation | ||
| Overfitting, strategies for handling them. Dropouts | ||
|
Keras & NN handson |
Keras Framework |
4 |
| CodingDemo n NN using Keras Framework and Native Python | ||
| Coding Problem in NN usin Keras | ||
| Test on NN, LogReg, LinReg | ||
|
Unsupervised Learning |
Unsupervised Learning | |
| K Means Cluster | ||
| Large Data sets, Batch Gradient Descent | ||
| Metrics for measuring performance – Accuracy, Errors, | ||
| Unbalanced data set: Precision, Recall, F1 Score | ||
| ROC/AUC |
| Feature Selection & Coding Demo | ||
| Summary concepts and How to choose a model | ||
| What next in ML? |
Trainer Summary:
Chandrashekar is an industry Veteran in Machine Learning, Deep Learning and NLP with 28+ years of experience in the Tech space. He conducts workshops on ML/DL/NLP solutions and consults with companies and implements ML solutions for them.