From 398863551448a1bec2c2f0a7982b625950694248 Mon Sep 17 00:00:00 2001 From: Peter Newell Date: Tue, 22 May 2018 20:25:43 +0800 Subject: [PATCH] Update README.md --- README.md | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/README.md b/README.md index ea3024a..a3fd1bd 100644 --- a/README.md +++ b/README.md @@ -21,8 +21,10 @@ Using a basic list of six standard Sci-kit Learn ensemble methods, we can explor A basic LSTM model built with Keras and using 20 input factors: open, high, low, volume, and technical indicators such as moving averages, a Stochastic Oscillator, Bollinger Bands, and others. The results are noisy and mixed, as it is unclear if more data helps or hurts the model from building meaningful connections between the data. + ![LSTMulti png](/graphs/LSTMulti Score 55.4103637890938.png) + This model could be improved in a number of ways. Increasing the number of layers, normalizing the data, reducing the amount of input data, and increasing the amount of learning data are the most obvious choices. ## Linear Models @@ -33,14 +35,20 @@ Linear regression models are natural candidates for time series analysis. Using ![Ridge png](/graphs/Ridge 0.8099881178871757.png) + *LinearRegression* ![LinearRegression png](/graphs/LinearRegression 0.8099495670315746.png) + + The Autoregressive Integrated Moving Average (ARIMA) is not exactly a machine learning algorithm, but a linear model used in econometric analysis that can be applied to financial markets to make predictions. A quick build of this model can also produce 78% accurate predictions on the test data. *ARIMA* + ![ARIMA png](/graphs/ARIMA 0.7895470127761371.png) + + ## Tensorflow NN This is a a FOREX adaptation of Sebastian Heinz's neural network for stocks from his Medium.com article ["A simple deep learning model for stock prediction using TensoFlow"](https://github.com/sebastianheinz/stockprediction).