This webpage contains an analysis of over 2000 mid to large size stocsk on the NYSE and NASDAQ using Random Forest modeling. A more technical description of this analysis is given at the end of the page.

Buy Recommendations

The following stocks were classified as gaining 5% in the next 5 days from today’s opening price. The stocks are sorted by accuracy of the fitted model.

Charts

StockCharts for up to the top 20 stocks are given below. The blue line is the EMA(5) indicator.

Technical Description

A Random Forest machine learning algorithm was used to fit an model of commonly used technical indicators to a classification of stocks in which a “1” indicates that the difference between the high price of the stock within the next 5 days and opening price of the stock tomorrow exceeded 5% of the opening price. The fitted model is then used to discover those stocks that are predicted to exceed this value based on yesterdays closing indicators. Specifically, the indictors used in model include Aroon, ADX, ROC, MACD(12,6,9), RSI, CMO, StochFastK, StochFastD, StochSlowD, SMI, StochRSIFastK, StochRSIFastD, StochRSISlowD, ChaikinVolatility, ChaikinMF, EMV, KST, BBands, MFI, Trix, UltimateOscillator, VHF, WPR, WilliamsAD, ADL(lagged), Volume(lagged), VolumeMA(lagged), OBV(lagged), ChaikinAD(lagged)

The model was fitted within each stock with available data back to the year 2007. The number of trees (ntree) and number of candidates at each split (mtry) were set to the default values of ntree=500 and mtry=floor(sqrt(x=29))=5. The model is not cross-validated, and hence the reported accuracy is the out-of-bag error for those predictions in which the model predicted a “1” or “buy”, i.e. the stock would exceed 5% in 5 days.