PI-660: Data Analytics |
(2 credits) |
1. | Pattern recognition: Introduction to pattern recognition and data mining, clustering vs. classification; applications; data handling and preprocessing, feature selection, normalization, dataset preparation: training, test, external; training of model; validation of model: internal validation, k-fold cross validation, external validation, y-randomization; applicability domain analysis, learning paradigms: supervised and unsupervised. |
2. | Machine learning algorithms for classification: k-NN, PNN, SVM. |
3. | Machine learning algorithms for clustering: Different distance functions and similarity measures, K-means clustering, single linkage and complete linkage clustering, hierarchical clustering, logic behind these algorithms. |
4. | Artificial intelligence: Overview on basic concepts and its application in Pharmacoinformatics. |
5. | Artificial neural network: Overview of biological neuro-system, mathematical models of neurons, ANN architecture, learning rules, ANN training algorithms-perceptions, training rules, delta, back propagation algorithm, multilayer perceptron model, applications of ANNs. |
6. | Genetic algorithms: An overview, GA in problem solving, implementation of GA, selection, mutations, crossover. |
7. | Fuzzy logic: Introduction to fuzzy logic, classical and fuzzy sets: overview of classical sets, membership function, fuzzy rule generation, operations on fuzzy sets: compliment, intersections, unions, combinations of operations, aggregation operations; application of fuzzy logic in medicine. |
8. | Expert systems: Expert systems (knowledge-based systems), expert system examples, expert system architectures, rule based expert systems, statistical systems, hybrid systems, non-monotonic expert systems, decision tree based expert systems. |
9. | R language: Introduction to R programming, functions, variables, data types, operators, data structures in R, objects, classes |
10. | Applications: Application of machine learning algorithms using R |
| |