BSCS 4-3: Elective 3 (Machine Learning) Midterm Exam - 5-18-2024

Which of the following is the primary objective of machine learning?

• To replace human workers with robots

• To allow software applications to become more accurate at predicting outcomes without being explicitly programmed to do so

• To perform complex calculations faster than humans

• To store and manage large datasets

Who designed the first neural network for computers and what is it commonly called?

• Alan Turing, Turing Machine

• Frank Rosenblatt, Perceptron Model

• Bernard Widrow, Adeline Model

• Gerald DeJonge, Explanation-based Learning

Which algorithm enables computers to perform basic pattern recognition and when was it introduced?

• Nearest Neighbor Algorithm, 1967

• Decision Tree Algorithm, 1975

• Support Vector Machine, 1980

• Backpropagation Algorithm, 1986

How did the approach to machine learning change during the 1990s? What were the key characteristics of this shift from earlier methods?

• From hardware-driven to software-driven approaches

• From statistical methods to heuristic methods

• From knowledge-driven to data-driven approaches

• From unsupervised to supervised learning approaches

A company is using machine learning to predict customer churn. They notice their model performs well on the training data but poorly on new data. Which term best describes this issue?

• Overfitting

• Underfitting

• Bias

• Variance

In the context of machine learning, how are "features" defined and why are they significant for building and training machine learning models? Which example best illustrates their role in the data analysis process?

• Features are the outputs generated by a machine learning model and are significant because they represent the predictions made by the model.

• Features are the individual measurable properties or characteristics of a phenomenon being observed, and they are significant because they serve as the input variables that the model uses to make predictions.

• Features are the raw data collected from various sources before any processing, and they are significant because they provide the initial information needed for any analysis.

• Features are the final adjusted weights used in the machine learning model, and they are significant because they determine how the input data is transformed into predictions.

In the context of machine learning, why is "big data" significant? How does the availability and utilization of large datasets impact the performance and capabilities of machine learning models?

• Big data reduces the computational complexity of machine learning algorithms

• Big data simplifies the process of feature engineering in machine learning models

• Big data allows machine learning models to be trained with less data, reducing the need for extensive data collection efforts

• Big data provides a diverse and extensive amount of information that enhances the learning process of machine learning models

When developing a machine learning model to predict house prices, if the model consistently predicts the same price regardless of the input features, what issue is this likely indicating about the model? What underlying problem could cause this behavior, and how might it be identified and resolved?

• The model is overfitting the training data

• The model is underfitting the training data

• The model is suffering from data leakage

• The model has a low learning rate

A healthcare company aims to leverage machine learning to predict disease outbreaks using historical data on disease incidence and various environmental factors. What type of machine learning task is the company likely undertaking? Consider the nature of the data and the objective of predicting future disease occurrences.

• Classification

• Clustering

• Regression

• Dimensionality Reduction

A social media platform wants to implement a recommendation system to personalize content for users based on their interests and preferences. Which term best describes the process of using machine learning to deliver personalized content recommendations to users?

• Prediction

• Overfitting

• Target (Label)

• Feature Vector