🎉 🎉 RAPIDMINER 9.10 IS OUT!!! 🎉🎉

Download the latest version helping analytics teams accelerate time-to-value for streaming and IIOT use cases.


stock trend prediction

fpaganelfpaganel Member Posts: 4 Contributor I
edited November 2018 in Help

Hi everyone,

I'm trying to develop a model that optimize buy/sell signals at the end of day, in order to obtain the best gain.

Inputs are techical indicators like RSI, MACD, Stochastic and candle patterns.

At the moment I have a training set of Apple's stock in excel (that I will share when I find out the way).

In the past I have tried to build a model considering following label:

  • 1 if the price variation is >0.2
  •  0 if the price variation is between -0.2 and 0.2
  • -1  if the price variation is <0.2

But the result accurancy is around 50% .

How can improve it?

Could you please help me? 


Thanks in advance




  • Thomas_OttThomas_Ott RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,760   Unicorn

    I would optimize your algorithm using the Grid Operator. You can get good success when you do that. 


    Also, visit my site on using RapidMiner for Stock Market related things. http://www.neuralmarkettrends.com/tutorials

Sign In or Register to comment.