| dc.description.abstract | In India, 40 % of startups close down within 12 months of operations. Of the 2281 startups that began operations in 2014, 997 have failed already. Several reasons are responsible for their failures including lack of sufficient funds, bad management, etc. Given the magnanimity of the factors that come into play, it gets difficult for a start-up, or even for the one that’s substantially funded, to zero in on the key areas to focus on. This work attempts to create a predictive model that goes on to decide the success or failure of a startup based on several factors at every given stage. Next, we study sample failures and abandonment cases like PepperTap, StayZilla, and the likes to test our model through the training data. A major point to note is that we don’t plan on providing the recipe for success, but to prevent a startup from failing by providing them with insights on what factors to focus on, depending on their stage.
The factors include seed funding time, funding amount, series A funding and many other factors that lead to the success or failure of a company at every milestone. We plan to create multiple models based on the data curated from various startup databases like TechCrunch, CrunchBase, Tracxn, CEB and lot more. Several data mining classification techniques are to be used on the data along with various data mining optimizations and validations. The analysis will comprise of techniques such as Regression, Factor Analysis, ADTrees and so on, depending on the need. The correctness of our models will be evaluated by examining the area under the ROC curve, precision and significance values. In short, we provide a model that helps a startup decide which factors, of many, they need to focus more on, to keep themselves afloat, or possibly hit the success mark. | en_US |
