Advanced Machine Learning with Deep Learning

• Introduction to Artificial Intelligence & Machine Learning
• Overview- AI Vs ML Vs Deep Learning
• Overview- Subfields of Artificial Intelligence- Robotics, ML, NLP, Computer Vision
• Applications of Machine Learning/AI
• Difference b/w AI & Programmed Machine
• R & R Studio Setup & Installation
• A quick tour of R-Studio – Variables, Install, Plot, help, console, repository
• Important Links to get datasets – Kaggle, data.gov, etc

• Classes & Objects
• Vector and List in R
• Hands-on

• Matrix & Factor in R
• Hands-on

• Dataframe in R
• Plotting using gggplot2 in R – Scatter plot, Box plot, Hist, Bar chart, etc
• N-Dimensional Array in R
• Table function in R
• Hands-on

• Statistics in R – Mean, Median, Mode, Range, Variance, SD, Inter Quartile
• Twitter- R Integration
• Get data from MySQL using R
• Get data from the website using the R
• Hands-on

• Steps involved in solving a Machine Learning Usecase
• Data preprocessing/preparation in R
• Missing data, Categorical data, Feature Scaling, Splitting data to test & train sets
• Hands-on with sample data

• Types of Machine Learning- Supervised & UnSupervised Machine Learning
• Supervised Learning – Regression & Classification
• UnSupervised Learning- Clustering
• Regression Algorithm- Simple Linear Regression
• UseCase: Create a Model to predict Salary from years of exp
• Classification Algorithm- K Nearest Neighbour
• UseCase: Create a Model to predict if a particular customer will purchase a product or not
• Hands-on with Sample data

• Clustering Algorithm- Kmeans
• Elbow Method in Kmeans to predict optimal no. of Clusters
• Clustering Algorithm- Hierarchical Clustering
• Dendograms in Hierarchical Clustering to Predict Optimal No. of Cluster
• UseCase: Using Kmeans & HC to extract patterns to analyse customer data based on spending score and income
• Hands-on with Sample data

• Logistics Regression
• UseCase: Create a Model to predict if a particular customer will purchase a product or not
• How to create and read the ROC curve
• How to check the accuracy of the Model using the Confusion Matrix
• Hands-on with Sample data

Day 10:

• Random Forest using Decision Trees
• Support Vector Machine for Classification
• UseCase: Create a Model using Random Forest & SVM to predict if a particular customer will purchase a product or not
• How to create and read the ROC curve
• How to check the accuracy of the Model using the Confusion Matrix
• Hands-on with Sample data

Day 11:

• Polynomial Regression
• UseCase: Create a Model to predict Salary from years of exp
• UseCase: Satellite Image Classification using Random Forest. Create a Model to identify/classify different types of land re.g, barren, forest, urban, river, etc. from a Satellite image
• Hands-on with Sample data

Day 12:

• Dimensionality Reduction
• Feature Selection Vs Feature Extraction
• Feature Selection using the Backward Elimination technique
• Feature Extraction using PCA
• Hands-on with Sample data
• How to tune/check the accuracy of the Model using P- Value, R Square, Adjusted R Square, CAP

Day 13:

• Overview of NLP/Text Mining
• Libraries in R for NLP/text mining – tm, Snowball, dplyr
• Bag of words using R
• Use Case: Restaurents Review System
• Sentiment Analaysis using R
• Use case: Analyse Twitter data for two teams to predict sentiments
• Hands-on with Sample data

Day 14:

• Overview of types of recommendation engines – Example E-commerce, Netflix etc
• Frequently bought items, User-Based Collaborative Filtering
• Libraries in R for recommendation – recommended lab
• Use Case: Analyse grocery store data to find out frequently bought together item
• Use Case: Analyse joke data to recommend the best jokes to users
• Hands-on with Sample data

Day 15:

• Time Series data analysis in R
• Components in time series – Trend, Seasonality
• Arima Model Vs ETS Model
• Use Case: Forecast Flight booking from Airline data
• Sentiment Analysis using R
• Hands-on with Sample data
• Deep Learning Introduction
• Limitations of ML and how Deep Learning comes to the rescue
• Biological Neural Network Vs Artificial Neural Network
• Popular Frameworks of Deep Learning – Tensorflow, Keras

Day 16:

• Understanding Deep Learning Terminologies – Input Layer, Hidden Layer, Output Layer, Activation Function, Cost Function, Back Propagation, Gradient Descent, Epoch, Learning Rate
• Install Keras (using tensorflow)
• Use Case: Create a model using ANN for Boston housing data

Day 17:

• Convolutional Neural Network
• Convolution, Polling, Flattening
• Use Case: Image classification using CNN
• Hands-on with Sample data

Day 18:

Case Study – Predict Customer Churn

Day 19:

Case Study – Canada Crime Analysis

Summary & QA