Scribd
Upload
5 views2 pages

Assignment Problem

The assignment involves predicting bearing failure using a Random Forest Classifier (RFC) with a provided dataset containing various input parameters and a failure type output. It consists of three parts: basic understanding of data preprocessing, intermediate analysis of model training and evaluation, and advanced exploration including hyperparameter tuning and feature importance analysis. Deliverables include a Jupyter Notebook, a PDF report summarizing the work, and a predictive maintenance system integration explanation.

Uploaded by

Sarayu K
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
5 views2 pages

Assignment Problem

The assignment involves predicting bearing failure using a Random Forest Classifier (RFC) with a provided dataset containing various input parameters and a failure type output. It consists of three parts: basic understanding of data preprocessing, intermediate analysis of model training and evaluation, and advanced exploration including hyperparameter tuning and feature importance analysis. Deliverables include a Jupyter Notebook, a PDF report summarizing the work, and a predictive maintenance system integration explanation.

Uploaded by

Sarayu K
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Assignment: Prediction of Bearing Failure using Random Forest Classifier

You are provided with a dataset of Bearing Failure, which contains multiple input
parameters related to bearing type and its operating conditions, and one output
column representing the failure type. Your task is to explore, preprocess, train, and
deploy a Random Forest Classifier (RFC) to predict the type of bearing failure.

Note: Each question carries 1 mark.

Part A: Basic Understanding


1. Import necessary libraries, load the dataset, and display: (a) number of samples
and features, (b) first and last five rows, (c) distribution of features and target
(failure) classes.
2. Check for missing values and explain how you would handle them. Also, generate
a correlation matrix plot for the input variables.
3. Identify categorical and numerical columns. Why is this distinction significant in
ML?
4. Encode categorical variables appropriately and explain why you prefer this
technique.
5. Scaled the data and split the dataset into training and testing sets (rigidly 80–20).
Part B: Intermediate Analysis
6. Train a baseline RFC using default parameters and report: accuracy, and
classification report.
7. Explain how an RFC makes predictions? Evaluate overfitting/underfitting using
training vs. testing accuracy. Suggest ways to reduce overfitting if observed.

Part C: Advanced Exploration


8. Perform hyperparameter tuning using GridSearchCV with parameters like
n_estimators, max_depth, min_samples_split, min_samples_leaf, etc. Report the
best parameters and accuracy, and prepare a classification report along with a
confusion matrix for the tuned model.
9. Plot the feature importance graph.
10. Make predictions for the following input parameters using the developed RFC,
interpret the results, and explain how this model could be integrated into a real-
time predictive maintenance system.
Type Air Temp Process Rotational Torque Tool Wear
(K) Temp (K) Speed (NM) (min)
(RPM)
H 290 310 1370 45.2 48
L 290 310 1450 50.3 9
M 290 310 1520 42.5 3

Deliverables
Students should submit:
(a) A well-documented Jupyter Notebook (.ipynb file) as well as a .pdf.
(b) A short report (2–3 pages) summarizing preprocessing, model performance before
& after tuning, and interpretation of results.

Note: Don’t use ChatGPT and other AI tools.

You might also like