Machine Learning Important Questions PDF
Document Details
Tags
Summary
These are important questions on machine learning. They cover topics such as hypothesis classes, different regression types, machine learning techniques, vector and matrix roles in ML, statistical measures, and performance evaluation. The questions also include data representation, supervised/unsupervised learning, reinforcement learning, statistical concepts, and exploratory data analysis.
Full Transcript
**II Year & I Semester ( AI -- A, B, C** **Machine Learning** **Important Question** 1. Explain Hypothesis class in detail 2. Explain the various types of regression 3. Explain the various types of Machine Learning technique 4. Mention the roles of vector and matrices in ML 5. Explain me...
**II Year & I Semester ( AI -- A, B, C** **Machine Learning** **Important Question** 1. Explain Hypothesis class in detail 2. Explain the various types of regression 3. Explain the various types of Machine Learning technique 4. Mention the roles of vector and matrices in ML 5. Explain mean, mode and median for different set of numbers (Example :list of natural numbers, list of random numbers) 6. Explain the various statistical measure to evaluate the performance of ML with example 7. Explain the various representation of input data set 8. Write short notes on a) Null Hypothesis and alternative hypothesis b) Supervised and Unsupervised Learning c) Reinforcement learning d)working Procedure of ML system 9. Describe the statistical concept in ML in detail 10. Illustrate exploratory data analysis with atleast 5 examples 11. Explain the various mathematical operations in vectors and matrices 12. Write short notes on Imbalanced data and compare the strategies used to balance the dataset. 13. Explain Multi class classification and how to learn multiple classes with one example 14. There are 500 documents out of which 390 are relevant documents. The system predicts 410 documents as relevant document out of which 40 are not actually relevant. Infer the confusion matrix from the above details and the related performance metrics. 15. Suppose I have 10000 emails in my mail box out of which 300 are spams. The spam detection system detects 150 mails as spam, out of which 50 are actually spams. What is the precision and recall of my spam detection system? 16. There are 500 documents out of which 390 are relevant documents. The system predicts 410 documents as relevant document out of which 40 are not actually relevant. Infer the confusion matrix from the above details and ROC by relating TPR and FPR 17. Explain how multi-class classification problem is solved 18. The confusion matrix for IRIS dataset is as follows +-------------+-------------+-------------+-------------+-------------+ | | Predicted | | | | | | Values | | | | +=============+=============+=============+=============+=============+ | Actual | | SETOSA | VERSICOLOR | VIRGINICA | | Values | | | | | +-------------+-------------+-------------+-------------+-------------+ | 19. | SETOSA | 11 | 2 | | +-------------+-------------+-------------+-------------+-------------+ | 20. | VERSICOLOR | 1 | 15 | | +-------------+-------------+-------------+-------------+-------------+ | 21. | VIRGINICA | 22. | 23. | | +-------------+-------------+-------------+-------------+-------------+ Find the TP, TN, FP and FN for finding SETOSA and VIRGINICA 24. Explain confusion matrix and related performance metrics with an example. 25. The confusion matrix for IRIS dataset is as follows +-------------+-------------+-------------+-------------+-------------+ | | Predicted | | | | | | Values | | | | +=============+=============+=============+=============+=============+ | Actual | | SETOSA | VERSICOLOR | VIRGINICA | | Values | | | | | +-------------+-------------+-------------+-------------+-------------+ | 26. | SETOSA | 11 | 2 | | +-------------+-------------+-------------+-------------+-------------+ | 27. | VERSICOLOR | 1 | 15 | | +-------------+-------------+-------------+-------------+-------------+ | 28. | VIRGINICA | 29. | 30. | | +-------------+-------------+-------------+-------------+-------------+ Find the accuracy, precision, recall and specificity in predicting VIRGINICA 31. Describe Probably Approximately Learning in detail 32. Compare the following 1. L1 regularization with L2 regularization 2. Binary Classifier and Multi-class classifier 3. Overfitting and Underfitting 4. Feature Selection and Feature Extraction 5. Dependence and Independence event 33. Define the following term a. Conditional probability b) Bias c) Variance d) Teachers noise e) L1 norm and L2 norm
