PREDICTION OF STOCKS USING MACHINE LEARNING METHODS

Retail trade or retail is a sale of certain goods to the end consumer or intermediary for further sale, which is implemented through trade in specially equipped premises, through personal sales, etc. Also, retail trade is a commodity exchange process aimed at meeting the demand of customers.

In addition, the retail sector currently occupies a leading position in terms of the intensity of development of the CIS countries economy. Excellent indicators have been achieved and many companies have reached a new level of trading. By about 2005, more than a dozen major retail chains had passed the billion-dollar milestone in terms of annual net revenue, and this is in dollars. The turnover of individual stores and retail facilities competed with some industrial enterprises with solid turnover and production bases.

Thus, we can claim that the sphere of trade affects the growth and development of related industries. The product promotion chain involves the participation of customers and their demand, as well as other participants in the process. Moreover, the development of trade requires sellers to pay more and more attention to working with the product range and inventory balances. Working with inventory and product balances is a main issue for many retailers. And the many companies needed to make sure that there is a sufficient quantity of goods in the warehouse. Another point is that, exclude overstocking, because this is also one of the problems of retailers with a high degree of accuracy is required to make decisions.

To sum up, making decisions in inventory management directly affects sales volumes, logistics costs, revenue, profit, and profitability. Inventory prediction is a necessary task to maintain an optimal level of inventory. I would like to note that the goal of the project / dissertation is to solve this problem using modern prediction methods based on machine learning technologies. The result is that in this way it is quite possible to analyze the dynamics of sales(consumer demand) thousands or even more products.

1. Jake VanderPlas. Python for Complex tasks: data science and machine learning. - St. Petersburg. January 2018

2. Wes McKinley. Python and Data Analysis. November 2015

3. Cholet Francois. Deep learning in Python. January 2018

4. Aurélien Géron. Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow, 2nd Edition September 2019.

5. Emmanuel Ameisen. Building Machine Learning Powered Applications. January 2020

6. http://pypi.python.org

7. http://www.enthought.com

8. http://pythonxy.googlecode.com

9. http://scikit-learn.org