Artificial intelligence can predict whether your company will become a new unicorn or collapse, scientists say predictions are 90 percent accurate
Artificial intelligence can predict how successful a new startup project will be with 90 percent accuracy, a new study using machine learning models that analyzed a huge number of startup companies has shown.
In most cases, startup companies end up proving to be a complete failure, more precisely, about 90 percent of them fail in the market. Between 10 and 22 percent of startups fail in the first year, a new study using artificial intelligence has shown.
According to CB Insights, as many as 42 percent of startup companies fail due to a misjudgment of market demand. According to Fundera, a whopping 82 percent of failed startups had a problem with cash flow from the start.
Given all this data showing the pattern by which startups fail, artificial intelligence can predict which of the new startups has the potential to become the new so-called unicorn.
A group of scientists, who published a study in the Journal of Finance and Data Science, managed to develop artificial intelligence that can do just that, that is, discover whether a particular startup will be the next biggest thing on the market or just another failed attempt to enter the market. .
The machine learning models used to predict that artificial intelligence have been analyzed by over a million companies to more accurately determine the success of a startup. If their tool proves to be completely accurate, more investors could help fund successful startups , while on the other hand reducing the risk of their investing in startups that ultimately turn out to be as much a waste of time as a waste of money.
The research results are promising, and artificial intelligence was able to successfully assess 90 percent of the startups involved in the study. So artificial intelligence has not yet fully predicted the development of all startups, but for a start, the results are more than good.
Artificial intelligence used data from the Crunchbase platform and compared the data with information on patents from the US Patent Office.
The scientists also noticed a problem with Crunchbase, due to its crowd source nature, that is, they noticed that some details about the companies involved in the study were actually missing.
However, the scientists recorded the missing data and subsequently used it to enter machine learning models. From this, they learned that the absence of some details is necessary to predict the future of the startup company.
By: Olivia J. - Gossip Whispers
