代写IOM103 ARTIFICIAL INTELLIGENCE IN BUSINESS 2nd SEMESTER 2023 / 24 MOCK EXAMINATION代做Statistics统计
- 首页 >> WebIOM103
2nd SEMESTER 2023 / 24 MOCK EXAMINATION
ARTIFICIAL INTELLIGENCE IN BUSINESS
Section 1: Single Choice Questions [20 points]
(a) (2 points) Which statement best describes heuristics in AI?
⃝ A) They provide precise and optimal solutions every time
⃝ B) They are algorithms that aim for acceptable solutions efficiently (Correct)
⃝ C) They are mainly used for database management
⃝ D) They always compute the longest path in a network
(b) (2 points) Which technique is not typically used in data preprocessing?
⃝ A) Principal Component Analysis ⃝ B) Feature Selection
⃝ C) Bootstrapping (Correct) ⃝ D) Feature Scaling
(c) (2 points) Linear regression is primarily used to:
⃝ A) Classify text into categories.
⃝ B) Predict values based on continuous variables. (Correct)
⃝ C) Cluster data into similar groups. ⃝ D) Enhance the contrast in images.
(d) (2 points) Logistic regression outputs: ⃝ A) A continuous outcome.
⃝ B) An integer count.
⃝ C) Probabilities between 0 and 1. (Correct)
⃝ D) Multiple classes simultaneously without modification.
(e) (2 points) What is a primary challenge in NLP? ⃝ A) Reducing the speed of processing.
⃝ B) Managing numeric data.
⃝ C) Understanding the context of conversations. (Correct) ⃝ D) Generating high-resolution images.
(f) (2 points) In reinforcement learning, what does the agent primarily learn from? ⃝ A) Predicted error gradients.
⃝ B) Interaction with the environment. (Correct) ⃝ C) Pre-labeled dataset.
⃝ D) Supervised feedback from a teacher.
(g) (2 points) Which is a famous example of a Large Language Model? ⃝ A) TensorFlow
⃝ B) GPT-3 (Correct) ⃝ C) Hadoop
⃝ D) Pytorch
(h) (2 points) What is a common activation function used in deep neural networks? ⃝ A) Fourier transform.
⃝ B) Relu (Correct) ⃝ C) Concatenate
⃝ D) Linear transform.
(i) (2 points) What is a potential impact of AI on future business strategies? ⃝ A) Reduced need for physical office spaces
⃝ B) Decreased reliance on human decision-making (Correct)
⃝ C) Increase in paper-based processes ⃝ D) Decrease in data usage
(j) (2 points) What is the fundamental building block of a deep neural network?
⃝ A) Decision trees
⃝ B) Neurons (Correct)
⃝ C) Linear regression models ⃝ D) Clusters
Section 2: Multiple Choice Questions (one or more correct answers) [20 points]
(a) (4 points) Which of the following are considered emerging AI technologies? ⃝ A) Genetic Algorithm
⃝ B) Linear Regression
⃝ C) Convolutional Neural Network (CNN) (Correct) ⃝ D) Federated Learning (Correct)
⃝ E) Blockchain
Section 3: Linear Regression [30 points]
Below is a small sample dataset to illustrate the types of data you would work with in predicting home prices based on square footage (x1 ), number of bedrooms (x2 ), and age of the home (x3 ). The target variable is the home’s selling price (y), measured in thousands of dollars.
Square Footage |
Bedrooms |
Age |
Price ($K) |
1500 |
3 |
10 |
300 |
2000 |
4 |
5 |
350 |
1800 |
3 |
20 |
280 |
2200 |
4 |
2 |
400 |
(a) (10 points) Write the input matrix X and output vector y, then use min-max normalization method to scale these two matrices. (keep 2 decimal places)
(b) (10 points) Build a linear regression model (θ) to predict home prices. (keep 2 decimal places)
(c) (10 points) Predict the house price when the size is 2300,bedrooms are 4, and age is 1.
Section 4: Logistic Regression [20 points]
A university wants to predict prospective students’ admission likelihood based on their GPA and GRE scores.
GPA |
GRE |
Admitted |
3.5 |
330 |
Yes |
3.0 |
310 |
No |
4.0 |
320 |
Yes |
(a) (10 points) Write the input matrix X and output vector y, then use the decimal scaling normaliza-
tion method to scale these two matrices to range [0-10]. (keep 2 decimal places)
(b) (10 points) Predict the admitted rate when GPA=2.8, GRE=320. (θ = [−12, 20, −16]T )
Section 5: Artificial Neural Networks [20 points]
A university wants to predict prospective students’ admission likelihood based on their GPA and intern- ship number. And build an ANN to model the data.
GPA |
Internship |
Admitted |
3.5 |
1 |
Yes |
3.0 |
3 |
No |
4.0 |
0 |
Yes |
ANN Structure:
• 2 input nodes;1 hidden layer with 3 nodes; 1 output node
• activation function in the hidden layer - ReLu - max(0, x)
in the output layer - Sigmoid -
W[2] = [2.6 -1 -0.5], b[2] = [-0.1] (a) (20 points) Predict the output matrix Y(ˆ) (keep 2 decimal places)
Section 6: Artificial Neural Networks [20 points]
A university wants to predict prospective students’ admission likelihood based on their GPA and intern- ship number. And build an ANN to model the data.
GPA |
Internship |
Admitted |
3.5 |
1 |
Yes |
3.0 |
3 |
No |
4.0 |
0 |
Yes |
ANN Structure:
• 2 input nodes;1 hidden layer with 3 nodes; 1 output node
• activation function in the hidden layer - ReLu - max(0, x), in the output layer - Sigmoid -
• loss function: J
W[2] = [ 1 1 1], b[2] = [0]
(a) (20 points) Update the W[1] , b[1] in 1 episode (α = 0. 1, keep 3 decimals)
Section 7: Open-ended Question [20 points]
Evaluate the application of Artificial Intelligence in enhancing financial analytics and decision-making within the financial services sector.
(a) (5 points) Identify specific AI technologies that can be integrated into financial analytics to enhance decision-making capabilities.
(b) (5 points) Discuss the advantages AI technologies offer in terms of improving the accuracy and speed of financial analysis.
(c) (5 points) Outline potential challenges financial institutions may face when incorporating AI into their decision-making processes, along with possible solutions.
(d) (5 points) Consider the ethical implications of using AI in financial decision-making, focusing on issues related to algorithmic bias, transparency, and customer privacy.