Aziz-s99/SVM-KNN-accruacy-and-effieciency-on-tunisieimmobilier.com-s-data-set

The accuracy and evaluation of both the KNN and SVM model, on the Tunisian real estate website: tunisieimmobilier.com ## Hello world! My name is Aziz Sghaïer. This is my first machine learning project that I made in 2021 back when I was a junior university student. I will explain in details everything you need to know about the data collection method, the dataset and the motives of the project ## The dataset The dataset contains a set of five variables and 118 observations. For each observation we have the title for the real estate ad on the Tunisie Immobilier website, it's location, price, surface and type. Type was an artificial variable made in order to distinguish the value and the category of each House or villa for sale. We had 4 types ranging from 1: small appartment, 2: fair appartment, 3: house and 4: Villa. ## Data collection method Every variable on the dataset except Type, was scrapped from Tunisie Immobilier's website using DataMiner browser extension. The coice of instances of the artificial variable Type wasn't generated randomly but was generated by a real estate broker after deep data observations and understanding. ## Project Motives I did this project with the intent to build two models that can predict the type or category of the real estate using the set of variables available, and to check the accuracy of both models and evaluate them. ## In conclusion This is a small project I made when I was still a college student and it got me really introduced to the world of data science and machine learning. Please feel free to use it for learning purposes, interpretations or even model implemntation.