https://github.com/chitranjan806/batch_performance_predictions_using_sensor_data

https://github.com/chitranjan806/batch_performance_predictions_using_sensor_data

Last synced: 4 days ago
JSON representation

• Host: GitHub
• URL: https://github.com/chitranjan806/batch_performance_predictions_using_sensor_data
• Owner: Chitranjan806
• Created: 2020-12-31T10:17:19.000Z (almost 4 years ago)
• Default Branch: master
• Last Pushed: 2024-08-15T08:20:44.000Z (3 months ago)
• Last Synced: 2024-08-15T09:48:21.900Z (3 months ago)
• Language: Jupyter Notebook
• Size: 2.73 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Batch Performance Predictions using Sensor Data

## Problem Statement

Using the given data of input parameters that are the 55 sensors of machines performing in batches at every time period i.e recorded at 7 instances we will find the statistical dependency. 

As this is a Regression problem to predict whether the machines are performing the best we have followed certain regression algorithms.

### Modelling

R2_Scores of every model for shuffled data, we obtained.

Using XGBRegressor >> 0.48


Using KNeighborsRegressor >> 0.42


Using RandomForestRegressor >> 0.38


Using GradientBoostingRegressor >> 0.36


Using Linear Regression >> 0.17


### The predictions as per XGBRegressor



### Results 

**The predictions with the test data were exported in a csv file, which was used by the submission portal to generate R2 scores**



### Outcomes

* Operational Efficiency:

By identifying performance trends over time, the model helps optimize machine operations, reducing downtime and improving overall efficiency. This leads to cost savings and increased production output.

* Anomaly Detection:

The model could be used to flag potential issues when predictions fall outside expected ranges, indicating machines that may require attention.