Mysuru, January 06:-MYRA School of Business hosted Dr Dmytro Karamshuk – Senior Data scientist at Skyscanner, UK for its 37th edition of InSight Special Lecture Series recently.
This 37thInSight Special Lecture Series was titled ‘Geo-Spotting – Mining Online Location-based Services for Optimal Retail Store Placement’. Senior Data Scientist, Skyscanner Dr Dmytra Karamshuk addressed the special lecture on the occasion.
According to Dr Karamshuk, Skyscanner helps its clients find the best location to begin a new business, or if they need to move their business to a new or different location for expansion purposes. For instance, if I want to open up a coffee shop somewhere in a town, Skyscanner will help me by not only looking at factors that will help, like by checking if it is located in a high-footfall area like next to a train station or a college campus as well as looking at various coffee shops in the area and using predictive analysis to find if that location is a good fit. Popularity of a venue also depends on the co-location with other venues according to Dr Karamshuk.
For example Starbucks popularity is highly correlated with train stations in New York. This is based on the mobility of passengers who visit Starbucks coming out of the railway station. To find the popular area one needs to define a radius with geographical coordinates, which depends on the number of venues in the area, heterogeneity of the venue types and percentage of competing venues.
Mobility features also depend on the factors like total number of check-ins, intensity of transition inside the area and intensity of transition outside the area based on Dr. Karamshuk’s research. Based on their potential and actual popularity, areas will be evaluated and ranked. Conducting regression will help in giving rank to the areas based on the significant factors of geographical and mobility features. Supervised learning method based on individual features helps in giving reliable and improved results for selecting an area for a retail store. –(KK)