1
$\begingroup$

Problem: Maintaining the products is a big task for a retailer. If we can estimate using AI to predict which products will sell the most, we can maintain sufficient stocks of the product without stocking excess inventory. This should be tracked with respect to time because I want to estimate the purchasing levels with respect to seasons.

We are building a AI model which can self learn with respect to purchasing data and provide an estimate to of which products should be purchased and in what volumes.

What approach should we take? As now we are currently thinking of using LSTM networks. Please give a proper approach and algorithm.

|improve this question
$\endgroup$
  • $\begingroup$ if anyone done this project or found in internet please keep the link $\endgroup$ – Harsha Satya Vardhan Vasamsett Jan 22 '19 at 12:42
  • $\begingroup$ Thanks for the question, it allows me to explain the difference between games and an AI which is playing the game. The first step is to model the problem into a business game. That means, there are products, consumers and transactions. Modeling the game has nothing to do with artificial intelligence but it's a pen&paper game or alternatively programmed in Javascript. And now, it is possible to analyze the game from the AI perspective. Which kind of decision is needed to win, can the outcome be predicted, is it possible to use a certain strategy? All these question have to do with core AI. $\endgroup$ – Manuel Rodriguez Jan 22 '19 at 13:37
  • $\begingroup$ @ManuelRodriguez can u elaborate it more please as i want to use python using keras library $\endgroup$ – Harsha Satya Vardhan Vasamsett Jan 22 '19 at 15:14
  • $\begingroup$ Keras is a deeplearning library which is used for playing Flappy Bird and similar games. At first, you will need a game. This can be created with an external programinng language or directly with Keras. Then it is called a forward model. It takes an user input and calculates the follow-up state. On top of the forward simulation the AI has to be created, similar like training keras to play pong. $\endgroup$ – Manuel Rodriguez Jan 22 '19 at 18:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.