geehaad/avrs-

Anomaly detection

Overview:

In this assignment we are given a bunch of time series data that are collected from “Qbot-Qdrone Leader-follower Demo” from AVRS research studio. The dataset is a 2 minutes experiment, contains the X and Y coordinates for the follower and the leader. In this project we use Pycaret library, which is an open-source machine learning library with different machine learning models.

We use 4 different unsupervised machine learning models:

1. SVM (binary)
2. KNN (unsupervised implementation)
3. PCA
4. DBSCAN

And then we plot the TSNE plots for each algorithm, and evaluate the algorithms by compare its output with the actual output, and compare the algorithms with respect to accuracy, precision, recall, and F1 scores.

Conclusion:

After applying the four algorithms, we used the labeled data to evaluate each model and compare its output with the actual output from the labeled dataset.