Machine Learning – Tensorflow

1. Introduction to Matplotlib

Different types of basic Matplotlib charts
Labels, titles and window buttons
Legends, bar charts, histograms, stack plots, pie chart
Loading data from a CSV & NumPy

2. Mathematics

Introduction to Metrics
Gradient descent and cost function
Hypothesis function
Regularisation in linear and logistic regression
Regression-underfitting and overfitting
Neural network in math

3. Pre Processing Data OR Outliers

Fact and dimensional analysis
Dimensionality reduction
Principal component analysis
Feature selection
Feature scaling

1.Introduction of Machine Learning

Motivation of machine learning
Use cases of machine learning
Future scope of machine learning
Real world domain using ML
Types of machine learning
Different tools/framework available for ML
Limitation of machine learning

2. Basics of Machine Learning

Understanding supervised & unsupervised learning.
Understanding Parametric and non Parametric Algorithm.
Regression and Classification.
Clustering techniques.
TF-IDF and vectorization techniques.

3. Machine Learning with Tensorflow

Motivation for deep learning
Case studies of deep learning
Revolution of Artificial Intelligence
Advantages of deep learning over machine learning
Real-life use cases of Deep Learning
Required linear algebra and statistics

4. Introduction of Artificial Neural Networks

Deep learning: A revolution in Artificial Intelligence
What is Neural Networks and historical background?
Neural networks vs. conventional computers
Basic structure of ANNs
How neural networks works?

5. Tensorflow Basics

What is Tensorflow?
TensorFlow in deep learning
Installation of tensorflow
Tensorflow code-basics
Constants, Placeholders, Variables, Scope, Graph,
Creating and saving a model
Re-Store a model
Running a simple ML algorithm on TensorFlow
Tensor Board: Graph visualization
Tensor Projector: Training pipe line visualization
MLP digit-Classifier using TensorFlow

6. Optimizers

What are optimization algorithms?
Derivatives of optimization algorithms
Gradient Descent
Stochastic gradient descent
Mini Batch Gradient Descent
Momentum
Adagrad
Adam

7. Activation Functions

What are activation functions?
Derivatives of activation functions
Step Activation Function
Sigmoid Activation function
Hyperbolic Tangent function- Tanh
ReLu- Rectified Linear units
Softmax Regression

8. Keras: A High Level API

Introduction to Keras
Keras vs TFLearn
Compose models in Keras
Sequential and functional composition
Predefined neural network layers
Using inception v3 predefined model
Batch normalization: Keras
Saving and Loading a model with Keras
Customizing the training process
Tensor-board: With Keras
Use-Case implementation with Keras

9. Tuning of Hyper-parameters

Train, Dev & Test sets
Bias & Variance
Regularization & Overfitting
Why regularization reduces overfitting?
L1/L2 & Dropout Regularization
Vanishing / Exploding gradients
Weight Initialization for Deep Networks
Numerical approximation of gradients
Normalizing activations in a network
Why does Batch Norm work?

10. Miscellaneous

What are computation graph?
derivatives with a computation graph
Broadcasting in Numpy/Tensorflow
Neural Network Representation
Computing a Neural Network's Output
Explanation for Vectorized Implementation
Parameters vs Hyper-parameters
Forward and Backward Propagation

1. Supervised Learning - Regression Problems

Simple Linear Regression
Linear regression with multiple variables
Linear Regression using Gradient Descent
Ridge and Lasso Regression Logistic Regression
Hands on with Regression problems using Tensorflow

2. Supervised Learning - Classification Problems

K-Nearest Neighbours
Decision Tree
Random Forests
Linear Discriminant Analysis
What is conditional probability?
Use Naive Bayes with python.

3. K-Nearest Neighbours

What is KNN classifier?
Calculate K in KNN?
Distance metrics: Euclidean and scaled Euclidean
Distance metrics: Cosine similarity
Complexity of brute force search
KD-tree representation
NN search with KD-trees
Complexity of NN search with KD-trees
Visualizing scaling behaviour of KD-trees

4. Support Vector Machines (SVM)

Learn the simple intuition behind SVM.
Maximal-Margin Classifier
Choose the right kernel for your SVM
Learn about RBF and linear kernels

5. Outliers

Remove outliers to improve the quality of your linear regression predictions
Remove the residuals on a real dataset and reimplement your regressor
Understanding of outliers and residuals on the Enron Email Corpus Understanding of outliers and residuals on the Enron Email Corpus

6. Dimensionality Reduction Techniques

Principle Component Analysis
TSNE
Multicollinearity
Hands on with Dimensionality reduction in Tensorflow

7. Decision Trees & Random Forests

Characteristics of Tree Models
Supervised Segmentation in Modeler
Defining Terms, Trees and Rules
What is CHAID
How does the CHAID algorithm work?
Building a quick CHAID model
Using Autoclassifer to run multiple CHAID settings
How does the C&RT algorithm work?
Surrogates for Missing Data
C&RT's Expert Settings
CHAID Expert Settings
What is Bagging and Boosting?

8. k-Means Algorithm

An unsupervised task
Hope for unsupervised learning, and some challenge cases
The k-means algorithm
k-means as coordinate descent
Smart initialization via k-means++
Other applications of clustering

1. Designing of ML Project

Orthogonalization
When to change dev/test sets and metrics
Why human-level performance?
Avoidable bias
Understanding human-level performance
Surpassing human-level performance
Improving your model performance

2. Projects

Handwritten digit classification
Recommender systems
Emotion/Sentiment Analysis

3. Test: Outliers

Carrying out error analysis
Cleaning up incorrectly label data
Build your first system quickly, then iterate
Training and testing on different distributions
Bias and Variance with mismatched data distributions
Addressing data mismatch
Transfer learning
Multi-task learning