Data Analytics in Healthcare

Organized by

Departments of

Biomedical Engineering, Computer Science & Engineering,

Computer Science & Applications

St.Peter's Institute of Higher Education and Research

(Deemed to be University U/S 3 of UGC Act 1956)

Avadi, Chennai - 600 054

Session 1: Introduction to Machine Learning

• Supervised learning

• Unsupervised learning

• Reinforcement learning

• Machine learning using python

• Machine learning using matlab

• ML in local machine vs ML in cloud

Session 2: Introduction to Mathematical concepts and setup installation

• White paper training for ML Algorithms

• Package installation

• Installing using distribution package tools

• Installing using python installer

• Installing from source

• Setting up python environment

• Scipy

Session3: Python- Hands on

• Introduction to Python

• Working with Variables in Python

• Python's Lists & Tuples

• Creating Python Functions

• Modules & Packages installation

Session 4: Machine learning algorithms-Hands on

• Linear regression with one variable

• Linear regression with two variable

• Naive Bayes classification

• Coding with python(without packages)

• Coding with python (with packages)

Session 5: Data Manipulation and visualization

• real time data gathering

• Visualizing the data using 2D models

• Visualizing the data using 3D models

• Overview of neural networks

• Experimenting ML with Google platform for Intelligence vehicle

• Technology(if camera is availed)

Session 6: K-Nearest Neighbours

• K Nearest Neighbors Classification

• K Nearest Neighbors Decision Boundary Value of 'K' Affects Decision Boundary

• Measurement of Distance in KNN

• - Euclidean Distance -Euclidean Distance (L2 Distance)

• - Manhattan Distance (L1 or City Block Distance)

• Scale is Important for Distance Measurement (Feature Scaling)

• Comparison of Feature Scaling Methods

• Feature Scaling: The Syntax

• Multiclass KNN Decision Boundary

• Regression with KNN

• Characteristics of a KNN Model

• K Nearest Neighbors: The Syntax

Session 7 : Model Generation For Medical Dataset

• K Value Affects Decision Boundary

• Choosing Between Different Complexities

• How Well Does the Model Generalize?

• Under fitting vs Over fitting

• Bias—Variance Trade-off - Training and Test Splits

• Using Training and Test Data

• Fitting Training and Test Data

• Train and Test Splitting: The Syntax

• Beyond a Single Test Set: Cross Validation (Training split, Test Split etc.)

• Model Complexity vs Error - Cross Validation