Skip to content

Latest commit

 

History

History
13 lines (10 loc) · 965 Bytes

File metadata and controls

13 lines (10 loc) · 965 Bytes

Predictive-Classification-Model-for-Airbnb-AI

• Gained insights into the New York City Airbnb rental properties and concluded the neighbourhoods with most attractive Airbnb rentals and the type of rental properties with most reviews. Furthermore, concluded the economic viability of the rentals with missing reviews through machine learning models such as k-NN, decision tree and gradient boosted tree (GBT) classifiers implemented via data science platform RapidMiner.

• Evaluated the performance of all the models and chose the model with best classification performance measure such as kappa, area under the curve. Furthermore, optimized the best performing model by changing the hyperparameters such as number of trees in GBT classifier for more precise prediction.

• Applied the trained model to a completely new data not utilized in the process of model building and reported the results comprising perceptible visualizations in a documented form.