Senz AIoT Savants - ML & C++

Numpy array arr is defined as follows : 

          arr = np.array([1,2,3,4,5,6,7])

Which one of the following returns the even numbers in the array ?

• print(arr(0: step = 2))

• print(arr(1: step = 2))

• print(arr[1::2])

• print(arr[::2])

Answer explanation

arr[::2] means to select every 2nd element from the array starting from the first element, i.e., [1, 3, 5, 7]. This works because the even numbers in the array are at positions that are multiples of 2, starting from the first element at position 0.

The other options are:

print(arr(0: step = 2)): This is not valid Python syntax. It should be arr[::2].

print(arr(1: step = 2)): This is not valid Python syntax. It should be arr[1::2].

print(arr[1::2]): This returns [2, 4, 6], which are the odd numbers in the array. This is because we start selecting elements from position 1, which is the second element, and then select every 2nd element from there.

A Python dataframe is defined according to the code shown in the picture. What is the data type (dtype) of column “Age” ?

• Int64

• String

• object

• Above snippet will give an error

Answer explanation

In the given DataFrame, the column "Age" has values of mixed data types - integers and a string ("hello"). When a column in a Pandas DataFrame has mixed data types, Pandas assigns the data type that can accommodate all the data in the column, which is usually the object data type.

In this case, since the "Age" column has a string value, Pandas assigns the data type of the column to object. If we try to perform numerical operations on this column, it will result in a TypeError due to the presence of the string value.

Which of the following is a method for reducing the dimensionality of a dataset?

• Normalization

• Standardization

• One-hot encoding

• Principal component analysis (PCA)

Answer explanation

PCA is a technique for reducing the dimensionality of a dataset by projecting it onto a lower-dimensional space while preserving as much of the variance in the data as possible. Normalization, standardization, and one-hot encoding are all preprocessing techniques that are not specifically designed for dimensionality reduction.

Which of the following evaluation metrics is commonly used for measuring the performance of object detection algorithms?

• Precision-Recall curve

• F1-score

• Mean Average Precision (mAP)

• Receiver Operating Characteristic (ROC) curve

Answer explanation

While Precision-Recall curve, F1-score, and Receiver Operating Characteristic (ROC) curve are also commonly used metrics for evaluating the performance of machine learning algorithms, they are more suitable for binary classification problems. Object detection is a multi-class classification problem, and mAP is specifically designed to evaluate the accuracy of object detection algorithms.

What is the output of the following code snippet?

• [1, 2, 3]

• [1, 4, 3]

• [1, 2, 4]

• [4, 2, 3]

Answer explanation

In this code, we create a list x with the values [1, 2, 3]. We then make a copy of this list and assign it to y. Next, we change the second element of x to 4. Finally, we print the value of y. Since y is a copy of x, changing x does not affect y, so the output will be [1, 2, 3].

Which of the following is NOT a common preprocessing step in machine learning?

• Normalization

• Standardization

• One-hot encoding

• Feature selection

Answer explanation

While feature selection is a common technique for reducing the number of features in a dataset, it is not considered a preprocessing step in the same way that normalization, standardization, and one-hot encoding are.

What is the output of the following C++ code snippet?

• 5 7

• 7 5

• 7 7

• Compilation Error

Answer explanation

This code initializes two integer variables x and y with the values 5 and 7 respectively. It then declares two integer pointers p1 and p2 and sets them to point to the memory addresses of x and y respectively.

The next line is where things get interesting: p1 = p2;. This line is assigning the value pointed to by p2 (which is y) to the value pointed to by p1 (which is x). So after this line executes, x will have the value 7, and y will still have the value 7.

Finally, the code prints out the values of x and y on the same line separated by a space using cout << x << " " << y << endl;. The output of this code will be: 7 7

In summary, the code swaps the values of x and y by using pointers, and then outputs the new values of both variables.

You work for a company that wants to predict which of its customers are likely to default on their loans. The company has a dataset that includes customer demographics, loan amount, loan type, credit score, and payment history.

What type of machine learning would be best suited to this scenario?

• Unsupervised Learning

• Supervised Learning

• Reinforcement Learning

• None of the above

Answer explanation

The scenario describes a classification problem where the goal is to predict whether a customer is likely to default on their loan or not. In supervised learning, the algorithm learns from labeled data (in this case, historical loan data) to make predictions on new, unlabeled data. In this scenario, the historical loan data can be used as the labeled data for training a supervised learning model such as a decision tree, logistic regression, or random forest to predict whether a new customer is likely to default on their loan or not. Unsupervised learning and reinforcement learning are not appropriate for this scenario as they do not involve labeled data and are used for different types of problems.

Which of the following is an advantage of using templates in C++?

• They allow for code reuse across different types

• They eliminate the need for error handling

• They can be used to define member functions outside of the class definition

• They improve runtime performance by minimizing function calls

Answer explanation

Templates are a powerful feature of C++ that allow for generic programming. By using templates, you can write code that can be used with different data types, without having to write separate implementations for each type. This allows for greater code reuse and flexibility, and can help reduce code duplication.

What is the output of the following C++ code?

• 1

• 3

• 5

• 7

Answer explanation

The expression *(arr + 2) is equivalent to arr[2], which accesses the third element of the arr array. Therefore, the output is 3.