Anomaly Detection Project
Budget: ₹600 – ₹1,500 INR
Dataset link - https://chitkara-hackathon.s3.ap-south-1.amazonaws.com/Chitkara_Anomaly_Detection.zip
We were assigned an anomaly detection project that I made. But my teacher said there are some problems in it.
1) We are doing kind of supervised learning but we need to find patterns. Basically instead of taking "login succesful" as a target we should take it as a feature. He said to take reference from here - https://scikit-learn.org/stable/modules/outlier_detection
2) Svc model takes too much time. Please see if anything can be done about that.
3) When I check for anomalies in my input data on web app, trained model should give reason too, why its an anomaly.
4) And put a feature that I can upload data file on web app and it can check the data and give anomalous records.
Project will be judged on these parameters-
1. Accuracy of anomaly detection and pattern recognition.
2. Efficiency and scalability of the model.
3. The ability of the model to predict the reason for the anomaly and the
attribute(s) responsible for it.
4. Create a UI and Backend is an optional bonus but above three points are
necessary
I can only give Rs 1000 through any UPI app as that's all I have (I cant use freelancer account due to issues). But I want this project to be the best and most impressive possible. Can anyone help? Please send a message. I am attaching my project as zip file.
We were assigned an anomaly detection project that I made. But my teacher said there are some problems in it.
1) We are doing kind of supervised learning but we need to find patterns. Basically instead of taking "login succesful" as a target we should take it as a feature. He said to take reference from here - https://scikit-learn.org/stable/modules/outlier_detection
2) Svc model takes too much time. Please see if anything can be done about that.
3) When I check for anomalies in my input data on web app, trained model should give reason too, why its an anomaly.
4) And put a feature that I can upload data file on web app and it can check the data and give anomalous records.
Project will be judged on these parameters-
1. Accuracy of anomaly detection and pattern recognition.
2. Efficiency and scalability of the model.
3. The ability of the model to predict the reason for the anomaly and the
attribute(s) responsible for it.
4. Create a UI and Backend is an optional bonus but above three points are
necessary
I can only give Rs 1000 through any UPI app as that's all I have (I cant use freelancer account due to issues). But I want this project to be the best and most impressive possible. Can anyone help? Please send a message. I am attaching my project as zip file.
Related categories:
Python
Algorithm
Machine Learning (ML)
Artificial Intelligence
Anomaly Detection