Refined_Lecture-1-Motivation & Applications.ppt
- 1. Machine Learning Techniques & its Applications (CAP535) Motivation & Applications Lecture-1 6BICTL School of Computing Data + Algorithms + Computing = Machine Learning
- 2. Data + Algorithms + Computing = Machine Learning
- 3. Data + Algorithms + Computing = Machine Learning Machine Learning- More Refined Definition
- 4. Data + Algorithms + Computing = Machine Learning Applications • Email Spam Detection • Image Recognition • Recommendation Systems • Autonomous Vehicles • Medical Diagnosis Classroom Exercise- Relate these applications to Tom Mitchell’s Definition of ML- 5 Minutes
- 5. Data + Algorithms + Computing = Machine Learning Example-1
- 6. Data + Algorithms + Computing = Machine Learning Example-2
- 7. Data + Algorithms + Computing = Machine Learning Example-3
- 8. Data + Algorithms + Computing = Machine Learning Example-4
- 9. Data + Algorithms + Computing = Machine Learning Example-5
- 10. Data + Algorithms + Computing = Machine Learning Traditional Programming Rule Based --🡪Follows explicit instruction Deterministic-🡪 Same input always produces same output Program Centric--🡪 Known Rules and Logic Limited Adaptability----🡪Cannot modify their behaviour (Unless you change the rule) Machine learning Programming Data Driven -🡪Learns from Data Non-Deterministic-🡪 Outputs can vary based on data Model Centric--🡪 Models that can learn from data(predict) Adaptability and Learning--🡪Refine their models with data
- 11. Data + Algorithms + Computing = Machine Learning Traditional Programming Vs Machine Learning Programming
- 12. Data + Algorithms + Computing = Machine Learning
- 13. Data + Algorithms + Computing = Machine Learning You can see hidden structure when the dimensions are reduced from 3 to 2 No labels and No feedback Labels and Feedback are given
- 14. Data + Algorithms + Computing = Machine Learning 5 Steps for approaching a Machine Learning application Problem • Define the problem to be solved • Collect the Data( Labelled or unlabelled) • Choose an Algorithm class • Choose an Optimization metric for learning the model • Choose a metric for evaluating the model
- 15. Data + Algorithms + Computing = Machine Learning What you will learn in this course ? • Linear Regression • Linear Classification • Regularization • Performance metrics • Probabilistic Generative and Discriminative models • PCA and SVD Algorithms for unsupervised mode • Support Vector Machines • Neural Networks • Deep Learning concepts • Convolutional Neural Networks • Hidden Markov Models • Recurrent Neural Networks-LSTM-GRU
- 16. Data + Algorithms + Computing = Machine Learning Evaluation: Internal Examination-Best two out of three CIA Exams –40 Marks(20 +20) + 10 Marks(Assignment) External Examination – End of semester - 50 Marks Question Pattern for CIA Part –A ( Answer any three our of four Questions)= 3 x 10 =30 Marks Part-B ( One compulsory Question) = 1 x 20= 20 Marks Semester Exam Pattern Part-A( Answer any four out of Six questions) = 4 x 20 = 80 Marks Part-B ( One compulsory Question) = 1 x 20 = 20 Marks
- 17. Data + Algorithms + Computing = Machine Learning Assignments: 5 + 5 = 10 Marks https://matlabacademy.mathworks.com/details/machine-learning-onramp/machinelearning https://matlabacademy.mathworks.com/details/deep-learning-onramp/deeplearning
- 18. Data + Algorithms + Computing = Machine Learning References: 1.Pattern Recognition and Machine Learning by Christopher M. Bishop 2.Introduction to Statistical Learning by Gareth James, Daniela Witten, Trevor Hastie, and Robert Tibshirani 3.Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow by Aurélien Géron 4.Gilbert Strang- Linear Algebra and its applications 5.Deep Learning by Ian Goodfellow, Yoshua Bengio, and Aaron Courville 6.Machine Learning: A Probabilistic Perspective by Kevin P. Murphy 7.Python Machine Learning by Sebastian Raschka and Vahid Mirjalili
- 19. Data + Algorithms + Computing = Machine Learning Thank you !