Machine Learning All Questions and Answers

Explain artificial intelligent agents versus machine learning algorithms.
Which issues should be kept in account to design a learning system? Mention the three features for the following learning task: Spam –nonspam Email classifier learning problem
Mention the importance and application of the following machine learning algorithms: Find-S, Candidate-Elimination, List-Then-Eliminate, ID3, Gradient Descent and Delta Perceptron Training Rule.
Describe training data, data mining, unsupervised and reinforcement learning.
State restriction and preference bias. Discuss over-fitting the data problem in decision tree learning.
Consider the House Price Prediction learning task, a linear regression problem. Now represent the appropriate training sample, hypothesis and the cost function. Also derive the general solution.
Develop the general solution to find the MAP hypothesis assuming any suitable concept leaning task. Apply Brute-force maximum-a-posteri(MAP) leaning algorithm for the same.
Discuss MAP and maximum-likelihood (ML) hypothesis.
Equate biological neural network to artificial neural network (ANN).
Design a two-input ANNs for the following Boolean operations with adjusted weights and coefficients: i) OR ii) NAND
Clearly mention the roles of weights and activation functions in ANN. Draw various types of activation functions used in ANN.
Mention the purposes of using the following machine learning algorithms: i) Bayes optimal classifier ii) GIBBS iii) Naïve Bayes classifier iv) Backpropagation
Show the flow diagram for the development process of an artificial intelligent agent using ANN. Which kind of problems cannot be solved by a single layer perceptron.
When should you use gradient descent and when normal equation method to solve a linear regression problem? Discuss in detail.

Describe briefly the face recognition procedure applying Backpropagation feed forward ANN concept.
Mention the significances of support vector machine (SVM) and k-nearest neighbor (KNN) classifier in machine learning.

Discuss genetic algorithm with its operators in detail.
Explain and design a learning system for tic-tac-toe game playing problem.
What is machine learning? How do machine learning tools help to develop an artificial intelligent agent?

In context of machine learning, define training data, data mining, supervised and reinforcement learning.
When and why will you use the following learning algorithms? i) FIND-S ii) CANDIDATE-ELIMINATION iii) LIST-THEN-ELIMINATE iv) ID3 v) Gradient Descent and Delta Perceptron Training Rule

Which issues should be kept in account to design a learning system? Mention the three features for the following learning task: Handwriting recognition learning problem
Define restriction bias and preference bias. When does over-fitting the data problem occur in decision tree learning?
Consider the House Price Prediction learning task, a linear regression problem. Now represent the appropriate training sample, hypothesis and the cost function. Also derive the general solution.
Develop the general solution to find the MAP hypothesis assuming any suitable concept leaning task. Use Brute – Force MAP leaning algorithm for the same.
Discuss maximum a posteriori and maximum likelihood hypothesis.
Compare between biological neural network and artificial neural network (ANN).
Design two distinct ANNs for the following operations with adjusted weights and coefficients: i) AND ii) NOR
Mention the role of weights and activation functions in ANN. Draw various types of activation functions used in ANN.
Show the flow diagram for the development process of an artificial intelligent agent using ANN. Which kind of problems cannot be solved by a single layer perceptron?
Mention the purpose of using the following machine learning algorithms: i) Bayes Optimal Classifier ii) GIBBS iii) Naïve Bayes Classifier iv) Backpropagation

When should you use gradient descent and when normal equation method to solve a linear regression problem? Discuss in details.
Discuss and design a learning system for checkers playing problem.
Mention the roles of support vector machine and k-nearest neighbor in machine learning.
What is genetic algorithm? Explain the pseudo-code of the algorithm with its in details.
Describe briefly the face recognition procedure applying feed forward ANN concept.
What are the purpose of studying machine learning in Computer Science and Engineering?
Define data mining, training data, supervised and unsupervised learning.
Write the FIND-S algorithm, also mention its drawbacks.
Which issues should be kept in mind to design a learning system?
When does over fitting occur in decision tree learning? What are the strategies to avoid over fitting the data?

Derive the general solution for the linear regression problem: House Price Prediction.
Define Restriction biases and Preferences biases.
Define Inductive bias. Give a model of inductive bias system for CANDIDATE-ELIMINATION Algorithm.
Define Machine learning and its application areas.
With proper example define the supervised and unsupervised learning.
Derive the gradient descent algorithm.
Define Decision Tree and write the ID3 algorithm.
With an example derive the expression of Entropy measurement.
Derive the expression of gradient descent algorithm for multi-variables.
Define the process of assurance that the gradient descent algorithm work properly.
Derive the expression of Logistic Regression, Linear Regression, Linear decision boundary and non-linear decision boundary.
Explain the gradient search to maximize Likelihood in a Neural Net.
Define the basic concept of principle Component analysis and briefly explain the PCA algorithm.
Explain the basic concept of K-Nearest Neighbor learning.
Define the notations of artificial neural network.
Derive the general solution of linear regression problem with gradient descent algorithm.
With an example explain the linear regression.
Derive the expression of linear regression for multi-variables.
Define genetic algorithm. Briefly explain the major operations of genetic algorithm.
With diagram explain the steps of genetic algorithm.

Explain the coding process to solve x – 2x + 3 = 0 using Genetic Algorithm.
Derive the expression of cost function for logistic regression.
With an example explain the multiclass classification problem with logistic regression.
Derive the expression of cost function for gradient descent algorithm.
Find out the maximum likelihood from Bayes theorem.
What is supervised learning? How does it differ from unsupervised learning?
When does overfitting occur in decision tree learning? What are the strategies to avoid overfitting the data.
In which type of problems the ANN learning technique is appropriate?
Describe biological and artificial neural network.
What is perceptron? Give a mathematical model f a perceptron.
Describe the various types of activation function. What are the roles of these activation functions.
What are the attempts you have to take to alleviate the problem of local minima in Backpropagation algorithm?
Describe briefly face recognition procedure applying neural networking concept.
What is the role of weights in ANNs?
What is hidden layer?