Contains my submission for the Spring Camp recruitment tasks for machine learning.
It has 2 segments:
-
KNN: I have applied KNN on the housing dataset of Kaggle and even provided a submission based on dropping features by RFE(Decision Tree Regressor) and later using KNN over chosen 14 features to predict the property price. The code uses a good deal of imports which are not all used but I had used them for comparison purposes and have willingly left them just to give an idea of the rigour of the solution.
-
Clustering: I have chosen 10 features (out of the 14 previously as the left out 4 were similar to one of the chosen ones) and applied clustering using K-Means on them with colour grading wherever necessary for distinction and better representation. These features as per data analysis would be sufficient to present a good enough prediction over property prices on the data set available on Kaggle.
The colaboratory notebook has its different cells dedicated to KNN Application and then the clustering of various features of the dataset.
- Gradio : I have left a link to my space on the google form itself. (Link: https://huggingface.co/spaces/Hsemih/PropPric/tree/main)