Досліджено практичні питання застосування нейромереж для розв’язування задач бінарної класифікації на малих обсягах вибірок. Розглянуто використання багатошарового персептрону та проблему оптимального вибору такого основного параметру нейромережі, як кількість нейронів на проміжному шарі. Досліджувалась задача прогнозування ймовірності успішності стартапів. Встановлено, що на даних вибірки малого об’єму, при неможливості отримати результат таким класичним методом, як побудова логістичної регресії, нейромережа у вигляді багатошарового персептрону дає високу точність класифікації. Для вибору оптимальних параметрів нейромережі потрібно проводити серії експериментів з метою встановлення взаємозв’язку між показниками ефективності нейромережі та зміною значення відповідного параметру, що надалі можливо апроксимувати функціональним зв’язком. Отримані залежності дають основу для вибору кращих параметрів нейромережі.

The paper is devoted to deepening the academic basics using forecasting modeling methods to determine the predictors of enterprises’ success. A startup as a form of entrepreneurship is important today due to the ability to maintain the sustainability of the economic system through a flexible response to challenges. The startup’s potential for receiving external, direct venture financing from other economic counterparties is important forits sustainable development and success. The empirical study puts forward two hypotheses. The first one is that successful startups have common features, which are the factors in obtaining venture financing, i.e. predictors of success. The second hypothesis is a continuation of the first one and requires testing the importance of information representation and clarity of future startup results among venture investors, in particular through the information available about the startup’s activity over the Internet. The empirical study is based on data sets about startups in Ukraine over the last decade. The simulation is performed with logit models developed by the authors. The calculation allows us to confirm the identification of factors of direct influence on the startup’s success according to the built models. The ability to obtain venture capital is one of the startup’s characteristics. The logit model is used as the research tool to determine the relevant factors for defining the positive decision of venture investors to provide startup funding. Predictors of obtaining external funding are identified and considered as the prerequisites for the startup’s success in general. According to the research results,the presence of previous investors, the startup’s profit orientation, the startup’s website, and availability of information about its activity in the social network are the important factors for receiving external financing by a startup. Thepaper argues that the startup’s focus on the public good without profit orientation does not stimulate venture investors. Two periods of the startup founding are singled out among the influence signs in deciding whether a startup will receive external financing: before 2014 and after it. The recognizability of a startup became the determining factor for venture financing after 2014 due to the information provided through the Internet. Until 2014, the relationship with large corporations’ clients had been the most important feature for a startup with external venture financing.

Predicting the success of a new venture has always been a topical issue for both investors and researchers. Nowadays, it has become even more relevant concerning start-ups-young innovative and technology enterprises aimed at scaling their businesses. The purpose of this study is to create a model for predicting start-ups’ success based on their descriptive characteristics. A model that connects such start-up features as the period from foundation to the first financing, the area of activity, type, and amount of the first financing round, business model, and applied technologies, with the start-up investment success, which refers to re-investment, has been developed using data from the Dealroom platform on statistics of start-ups activity and their description. The final sample included 123 start-ups that are founded or operate in Ukraine. Three machine learning algorithms are compared: Logistic Regression, Decision Tree, and Random Forest. Acceptable results were obtained in terms of Accuracy, Sensitivity, and F-score, despite the limited data. The best model concerning start-up success prediction is determined by a Decision Tree, with an average effectiveness of 61%, 55%, and 52%, respectively. The AUC level for the Decision Tree achieved 58%, which is lower than the Logistic Regression and Random Forest scores (65%). But the last models had done so well by better predicting start-up failures, while more practical is the ability to predict their success. All models showed an acceptable level of AUC to confirm with confidence their effectiveness. The decision support system for the investment object can be helpful for entrepreneurs, venture analysts, or politicians who can use the built models to predict the success of a start-up. This forecast, in turn, can be used to drive better investment decisions and develop relevant economic policies to improve the overall start-up ecosystem