Post Graduate Programme in Data Science for Climate & Health 

offered by Work Integrated Learning Programmes

Course 1
Regression

Regression as a type of supervised learning technique where the target attribute is a continuous variable; regression models from theoretical and implementation perspectives

Course 2
Feature Engineering

Feature Engineering as a step to develop and improve performance of Machine Learning models; Data wrangling techniques that help transforming the raw data to an appropriate form for learning algorithms; Data preprocessing techniques such as normalization, discretization, feature subset selection etc. and dimension reduction techniques such as PCA

Course 3
Classification

Classification is a type of supervised learning techniques where the target attribute takes discrete values; Three types of techniques to solve classification problems – discriminant function, generative, and probabilistic discriminative approaches

Course 4
Unsupervised Learning and Association Rule Mining

The course focuses on finding natural groups or clusters that are present in the data. The course will cover lustering algorithms like K-means, Hierarchical & DBSCAN algorithms, Hidden Markov Models for time series prediction, and market basket analysis to generate the interesting rules from a transactional database

Course 5
Data Science for Climate Change

Evolution (long-term climate data time series analysis, simple statistical models etc), current extent (spatial visualization, new data collection techniques such as AWS, satellite based platforms and citizen science based data collection, its assimilation) and future projections (regional climate modelling, climate data downscaling, and bias correction using deep learning and other DS tools) of the climate change at global, regional and local scales; Solution concepts such as GHG inventory, mitigation pathways (from simple statistical models to complex integrated Assessment model – IAMs); theories and practical case-studies; social aspects of data collection, selection and use (biases, distortions, and blindspots, and the role governance and ethics)

Course 6
Data Science for Health

Need for ML in healthcare, Real world applications and examples; Different data types available from healthcare systems (EMR, population, surveillance etc.); Handling of unstructured data (medical images, clinical text, Biomedical signals); ML techniques for health data; Deployment of AI models in clinical workflows; Challenges in clinical ML - data challenges, interpretability; Ethical and regulatory issues for AI in healthcare - bias, fairness, privacy and security considerations

Capstone Project

Real life problems encompassing a typical data science pipeline obtained from organizations/third party vendors; Jointly mentored by the industry experts and faculty; Comparative study of the relevant techniques covered in the VII-50 course; Presenting the results in the required format; Fortnightly review of progress of the project