Indories..

Being Indorie we are proud of calling it a foodie city that has made up to most of its culture.

Introduction

There are many restaurants in Indore city and many more are being built but it is quite difficult to find out which locations are perfect to open a new food brand and also with this increasing number of places we are not able to figure where to go what to eat this all takes up alot of time to select according to our mood and atlast we are left with confusion.

The goal is helping people to make more inclusive and thoughtful choices when they are going to have lunch, dinner or brunch with friends, family or colleagues.

The objective of this is to analyse and select the best locations in Indore city to open a new Restaurant or finding best places to visit. Using data science methodology and machine learning techniques like k-means clustering, data visualization.

Data

We have required two types of data.

1. Indore Resturants data that contains list Locality, Resturant name,Rating along with their latitude and longitude.

• Data Source: Zomato kaggel dataset
• Description : This data set contains the required information. And we will use this data set to explore various locality of Indore.

2. Nearby places in each locality of Indore.

• Data source : Fousquare API
• Description : By using this api we will get all the venues in each neighborhood.

After converting the selected data into a dataframe, we export and save in csv format. Finally, we visualize the data for better comprehension. Here a map showing all the food places in indore.

Food places in Indore

Approach

• Collect the Indore data from Zomato kaggel dataset.
• Using FourSquare API we will find all venues for each neighborhood.
• Filter out all venues that are nearby by locality.
• Using aggregative rating for each resturant to find the best places.
• Visualize the Ranking of neighborhoods using folium library(python).

Analysis

1. Top 10 restaurant of indore using aggregate ratings from the zomato data set are:

Top 10 Restaurants 
1.Kebabsville - Sayaji Hotel
2.Joshi Dahi Bada House
3.The Cube - Hotel Effotel
4.Cravings - Sayaji Hotel
5.Mustang - Hotel Somdeep  Palace
6.Vijay Chaat House
7.Naturals Ice Cream
8.Chang's Legacy
9.Ni Hao - Radisson Blu Hotel
10.Simi's

2. Number of Restaurants based on the type of cuisine available:

Most available cuisines are — North Indian,Chinese,Fastfood.

3. Clustering the data into 7 cluster:

Clustering is done here on the basis of 1st Most Common Venue.

Cluster map of indore

Here different color show clusters sequentially from 0 to 6.

After we have segregated the dataset in clusters, we analyze them to gain a better insight. We plot bar graphs for each neighborhood to get a better insight on the distribution paradigm of restaurants in each cluster.

Results:

1. Cluster-0: Include Fastfood Restaurant and Indian Restaurant mainly.

There are 1692 places where 1st most common choice is Fast Food Restaurants.

This list can be accesed in my notebook.

2. Cluster 1: Includes Indian Restaurant Majorly.

There are 453 places where 1st most common choice is Indian Restaurants.

Similar works for remaining clusters…

In this way we can check on different locality and see which type of cuisine is popular there and how many such restuarant are available to check on the market and possibilites to open up a new place.

links

1. https://www.kaggle.com/sakshiwork/zomato-indore-data
2. https://github.com/sakshi13-m/capstone-project