SimpleRNN realizes stock forecast

Please check the previous articles for the principle. 1. Data source   SH600519.csv is the daily k-line data of sh600519 Guizhou Maotai downloaded by tushare module. In this example, only its C-column data is used (as shown in the figure):   use the opening price for 60 consecutive days to predict the opening price on ...

Posted by phpnewbie25 on Thu, 19 May 2022 23:01:07 +0300

[from the official case study framework Keras] seq2seq based on character LSTM

[from the official case study framework Keras] seq2seq based on character LSTM Keras official case link Tensorflow official case link Paddle official case link Pytoch official case link Note: this series only helps you to quickly understand and learn, and can independently use the relevant framework for in-depth learning research. Please ...

Posted by sarah on Sun, 08 May 2022 07:29:41 +0300

[Keras] TensorFlow distributed training

When we have a large number of computing resources, we can make full use of these computing resources by using appropriate distributed strategies, so as to greatly reduce the time of model training. For different usage scenarios, TensorFlow is in TF distribute. Strategy provides us with several distributed strategies, which enable us to train t ...

Posted by Mercenary on Thu, 05 May 2022 15:33:08 +0300

Using scikit learn in Keras

Created on: April 15, 2022 Revised on: April 15, 2022 1. Introduction Keras focuses on deep learning, not all machine learning Scikit learn machine learning library based on SciPy The Keras class library provides a Wrapper class for the deep learning model, which packages Keras's deep learning model into a classification or regression m ...

Posted by kaser on Fri, 15 Apr 2022 13:32:00 +0300

Simple and beautiful in-depth learning package - bert4keras

Novice friendly bert4keras During the internship in goose factory, follow Su Shen( Scientific space )The blog inspired idea and successfully improved an online model. The idea output and experimental progress are largely due to the design of Su Shen bert4keras It is clear, lightweight and based on keras. It can implement Bert concisely. At the ...

Posted by ess14 on Fri, 01 Apr 2022 18:55:00 +0300