Scribd
Upload
41 views

AML PRG Assign I

This programming assignment involves analyzing a student data set to: 1) Perform exploratory data analysis including descriptive statistics and data visualization. 2) Frame a problem statement by establishing a relationship between variables based on the analysis. 3) Suggest and evaluate a regression model to predict a dependent variable from independent variables. Students are asked to save their findings in a Word file and Python notebook, and submit both files zipped together. Guidelines provide details on required tasks like data cleaning, correlation analysis, and evaluating the regression model. Presentation and insights are also important parts of the assignment.

Uploaded by

Padma
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
41 views

AML PRG Assign I

This programming assignment involves analyzing a student data set to: 1) Perform exploratory data analysis including descriptive statistics and data visualization. 2) Frame a problem statement by establishing a relationship between variables based on the analysis. 3) Suggest and evaluate a regression model to predict a dependent variable from independent variables. Students are asked to save their findings in a Word file and Python notebook, and submit both files zipped together. Guidelines provide details on required tasks like data cleaning, correlation analysis, and evaluating the regression model. Presentation and insights are also important parts of the assignment.

Uploaded by

Padma
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Birla Institute of Technology & Science, Pilani

Work Integrated Learning Programmes Division


Second Semester 2022-2023

Programming Assignment-I
(EC-2 Regular)

Course No. : SE ZG568/SS ZG568


Course Title : Applied Machine Learning
Weightage : 12%
Duration : March 20-30, 2023
Course Instructor : Dr Bharathi R

Assignment Objective: To analyze a given data set. Perform Exploratory Data


Analysis. Suggest a regression model.

Dataset: student.csv
The final deliverables of the Programming Assignment-I are
i) a word file documenting all the findings of every stage
ii) Python code in ipynb format
Save both files in a folder, zip and upload.

Tools and Techniques


Python libraries for data analysis.
(NumPy,SciPy,Matplotlib,Pandas,ScikitLearn,Statsmodels,Seaborn,Bokeh,Blaze,Scrapy,Req
uests,BeautifulSoup)

Sample Exploratory Data Analysis Case Studies


https://www.kaggle.com/c/house-prices-advanced-regression-techniques
http://ucanalytics.com/blogs/exploratory-data-analysis-retail-case-study-example-part-3/

Programming Assignment -I Guidelines


These are the guidelines and questions that you are expected to answer. The student will have to
analyze the data that he/she has been given and come up with meaningful insights for the given
dataset. They have to decide a problem statement based on the dataset that they have received after
they have performed descriptive statistics and data visualization. The steps that have to be taken are
explained below.

1) Descriptive Statistics
Data given in the dataset has to be understood and every feature must be explained by the student.
The datatypes present in the dataset must be found out. The measures of central tendency should
be found and explained. Based on these values, there should be a few critical insights made that
would then lead to their problem statement. Data cleaning should also be performed by
suggesting appropriate techniques to handle missing data and outliers.

Note: Exploratory Data Analysis (EDA) is used to tackle specific tasks such as:
i. Spotting mistakes and missing data;
ii. Mapping out the underlying structure of the data;
iii. Identifying the most important variables;
iv. Listing anomalies and outliers;

2) Data Visualization
Data should be visualized using the various types of charts and graphs that the student has learnt.
Utilizing these visualizations, there should be insights from every visualization that is submitted
and they should help frame the problem statement that is intended to be solved.

3) Framing problem statement


The problem statement should be based on the numbers and visualizations that had been done so
far. The problem statement should aim to establish a relation between variables. The problem
statement should meet the following criteria
 Can be proven true or false
 Should be detailed and mention both the dependent and independent variable.
This has to be approved and written down

4) Coming up with correlation analysis


Correlation Analysis must be done to find out how the variables are related and how the regression
model could be made.
5) Regression Model
A regression model should be established between the selected variables and should be used to
predict the values of the dependent variable.
6) Evaluating said regression model
The regression model should be evaluated using the error scores and attempts should be made to
increase the accuracy of the model and decrease the error of the model by using various variables.

Presentation

 Presentation is key. Ensure that your notebook is capable of explaining your insights and
visualizations by itself. Section your questions and emphasize your results. Do not hide your
final result in a sea of code or debugging cells.

Examples:
 If your question is on data cleaning, highlight the rows which need to be cleaned and
show the results of your data cleaning before and after it has been applied on those
rows.

 If your question asks you to prove a statement using visualizations, ensure that you
actually have a concluding statement after your graphs. Do not leave the conclusion
unstated after visualizing the data in your notebook.

 It is recommended to have short bullet points explaining what you have done before each task,
especially for non-visualization tasks. This will help us understand your approach to the
problem and can help with partial marks even if you are unable to solve the entire question.
 Prioritise interpretability over design. While it is encouraged to have visually appealing graphs,
make sure that you do not lose interpretability of the data in the pursuit of aesthetic
visualizations.

Insights

 The last section of your report will have to be dedicated to an out of the box pursuit. If you think
you have a better way of cleaning the dataset or visualizing a question, or if you believe that you
have noticed an interesting insight that can be cleaned from the data, add them at the end of your
notebook and elaborate why you think you’re right in your report or notebook and make sure
you mention it in your recorded video. This carries weightage to your final scores.

You might also like