ix$.$ What is the main difference between the linear regression and logistic regression?

x$.$ Which of the following Al algorithm has the guaranty to get global optimization?

$1-$Na$$ve bayes$2-$perceptron$)3-$logistic REG$)4-$SVM$)5-$K$-$mean$)$

xi$.$ Can the machine learn the features, if yes, in what NN this could happen? if no$,$ why?

xii. What is the most important theorem in DNN$?$ Give brief description of it$.$

xiii. List all the parameters that could be designed or changed in DNN$.$

xiv. Clustering is one of ML algorithms, is it supervised learning? what about the data is it

labeled? If is not labeled, what technique could overcome this issue?

xv$.$ SVM is a linear classifier, what is the major difference between SVM and linear perceptron?

xvi. SVM is one of the best methods that use generalization approach, provide evidence for this

claim.

xvii. What about if SVM faces non$-$linearly superable data, what is the solution, explain the

mechanism of this solution.

xviii. When SVM is not preferable to use?

xix. Let us consider another problem that SVM suffers from, which is when some instance falls

within the margin, can kernel function solves this matter, if yes how. If no$,$ how to overcome

this problem?

xx$.$ Is K$-$mean algorithm affected by the initialization process if yes, what is the solution then?