Fundamentals of Data Science: what's likely to come

Define entropy. How is it used with decision tree?

Describe and explain with examples the various types of machine learning methods (Supervised, Unsupervised, and Reinforcement).

OR

Explain with examples and highlight the relationship between Artificial Intelligence and Machine Learning.

You want to identify global weather patterns that may have been affected by climate change. To do so, you want to use machine learning algorithms to find patterns that would otherwise be imperceptible to a human meteorologist. Discuss what machine learning method (supervised, unsupervised, reinforcement) would you use and why.

The following table represents a dataset of 10 objects with attributes Color, Type, Origin and the "class", whether the customer who bought was satisfied or not.

Use ID3 algorithm to find the attribute with maximum information gain.

OR

What is multi-layer neural network? How is learning done in neural networks? Explain backpropagation algorithm. [1 + 1 + 4]

Explain node and weights in neural networks. Consider following Neural Network and compute its output considering sigmoid as activation function for all layers. Weights of synaptic links are provided above each link.

Input: $X1 = 2$, $X2 = 3$. Feedforward net: input nodes feed hidden nodes 1 and 2, which feed nodes 3 and 4, which feed output node 5. Weights: $X1 \to$ node 1 = 0.8, $X1 \to$ node 2 = 0.4, $X2 \to$ node 1 = 1, $X2 \to$ node 2 = 0.6, node 1 $\to$ node 3 = 1.2, node 1 $\to$ node 4 = 0.4, node 2 $\to$ node 3 = 0.7, node 2 $\to$ node 4 = 0.5, node 3 $\to$ node 5 = 1.5, node 4 $\to$ node 5 = 0.5.

OR

What is multi-layer neural network? How is learning done in neural networks? Explain backpropagation algorithm.

How does machine learning differs from traditional learning? Explain the various type of machine learning techniques.

Consider a dataset representing whether students passed an exam based on three features: Study Hours (Low, Medium, High), Previous Grades (Low, Medium, High), and Tutoring (Yes or No). The target variable is Exam Result (Pass or Fail).

Using the ID3 algorithm, calculate the information gain for each feature (Study Hours, Previous Grades, Tutoring) and determine which feature should be chosen as the root node for the decision tree.

OR

Consider a dataset containing the coordinates of 8 points in a two-dimensional space:

• Point 1: (2, 3)
• Point 2: (3, 4)
• Point 3: (3, 5)
• Point 4: (4, 6)
• Point 5: (7, 8)
• Point 6: (8, 7)
• Point 7: (9, 8)
• Point 8: (10, 9)

Apply the K-Means algorithm to cluster these points into 3 clusters.

The following table presents a dataset of 11 objects with attributes Color, Type, Origin and the 'class' whether the customer who bought was satisfied or not.

Now classify a new object using Naïve Bayes classifier with the following properties: Color = Red, Origin = Imported and Type = Formal

OR

What is clustering technique? Divide the data points $\{(2,10), (2,5), (8,4), (5,8), (7,5), (6,4)\}$ into two clusters using K-Means Clustering technique. [1+5]

Compute the output of following neural network using sigmoid activation function. Weights of synaptic links are provided above each link.

Input: $X1 = 2$, $X2 = 3$. Network is a feedforward net with input nodes feeding hidden nodes 1 and 2, which feed nodes 3 and 4, which feed output node 5. Weights: $X1 \to$ node 1 = 0.8, $X1 \to$ node 2 = 0.2, $X2 \to$ node 1 = 0.4, $X2 \to$ node 2 = 0.6, node 1 $\to$ node 3 = 1.2, node 1 $\to$ node 4 = 0.4, node 2 $\to$ node 3 = 0.7, node 2 $\to$ node 4 = 0.5, node 3 $\to$ node 5 = 1.5, node 4 $\to$ node 5 = 0.5.

What do you mean by support vectors? Explain with appropriate example of your own how supports vectors are useful in machine learning?

Explain neural network along with forward propagation and back propagation.

OR

Differentiate between Regression and Classification in supervised learning.

Apply Naïve bayes algorithm to decide if Married Female with salary of 42000 is probable to have illness or not based on data given below:

OR

Explain the forward propagation and backward propagation of neural networks.

Elaborate the concept behind Naïve Bayes algorithm for classification task.

Please answer any ONE of the following:

(a) Explain what is Naïve Bayes. Describe its common use cases with examples.

OR

(b) Explain Decision Trees. Describe its common use cases with examples.

Explain how classification is done by SVM classifier.

Discuss on machine learning and its types.

What is deep learning? How is it similar or different from neural network.

What do you mean by backpropagation? Conceptually, how do they differ from forward propagation in neural network?

How machine learning is different from traditional programming? What actually machine learns through machine learning? Justify your answer.

What do you mean by deep learning? How is it similar/different to Neural Network?

List the differences between supervised and unsupervised machine learning methods including examples.

Define and briefly describe neural networks. List a few example use-cases where neural networks can be used.

What are decision trees? What does impurity of a node mean in context of decision tree?

What kind of problem can a Decision Tree solve? Explain with an example.

Define Support Vector Machines. Describe briefly how SVMs are used for classification.