EXIN AIMLA Certification Sample Questions

01. Dale is wishing to develop an application of machine learning that is able to sort different types of user requests that have been manually input into a system. Due to the volume of user requests it is often time-consuming for an individual to read through each request. It is not always easy to quantify or prioritize requests based on how many times the same type of request is made.
Dale has observed that there are specific words that regularly feature in certain types of user requests that could be used to identify them. He would therefore like a machine learning model to read through each request and sort them into defined categories based on whether these specific words feature.
What type of approach should he use to solve the problem?
a) Regression
b) Classification
c) Clustering
d) Reinforcement

02. Roisin is training her machine learning model using unlabeled data and no training data. What type of type of approach would she use?
a) Semi-supervised learning
b) Reinforcement learning
c) Supervised learning
d) Unsupervised learning

03. Which of the following two frameworks can be used to develop machine leaning models?
Please remember to choose 2 answers.
a) Google analytics
b) TensorFlow
c) Scikit-Learn
d) Minecraft

04. Satpal has been developing an application (App) that can be used to order food from different restaurants and have it delivered straight to your home.
As part of the functionality, he has built in a machine learning model that uses regression to provide the user with an 'expected delivery time' for the food based on the time of day, the distance between the restaurant to the intended location, and the average delivery time.
The algorithm being used has been configured to compare the two variables 'time of day' and 'average delivery time' in order to make its prediction. When testing the App, he has found that the predicted 'expected delivery time' seems incredibly long based on his location to the restaurant.
What is probably the issue?
a) The choice of algorithm
b) The average delivery time
c) The average recorded speed of the driver
d) The variables being compared in the data

05. Which two of the following problems can be solved through classification?
Please remember to choose 2 answers.
a) Identifying an image based on specific features in the data
b) Grouping sets of unlabeled data to identify different customer segments
c) Sorting emails into 'received' and 'spam'
d) Making predictions on the number of cases of a virus in a particular area

06. What is the correct order of steps in the machine learning process?
a) 1. Problem identification
2. Data selection
3. Data pre-processing
4. Training
5. Testing
b) 1. Problem identification
2. Data pre-processing
3. Data selection
4. Testing
5. Training
c) 1. Problem identification
2. Data pre-processing
3. Data selection
4. Training
5. Testing
d) 1. Problem identification
2. Training
3. Data selection
4. Data pre-processing
5. Testing

07. What is the purpose of data pre-processing?
Please remember to choose 2 answers.
a) To clean the data to ensure it is suitable for training a machine learning model
b) To identify features or target values in the data that can be used to create training data
c) To identify the types of data required to solve the problem
d) To present the data using a graph or chart

08. Machine learning can be used to sort unlabeled data into groups. What is this known as?
a) Classification
b) Prediction
c) Grouping
d) Clustering

09. Which two of the following languages are commonly used in machine learning?
Please remember to choose 2 answers.
a) SQL
b) MATLAB
c) Python
d) CSS

10. When would you use unsupervised learning?
a) When we have sets of labeled data
b) When our algorithm does not require our input
c) When we are too busy to oversee the process
d) When we have sets of unlabeled data