I am supposed to predict column E but I cannot figure out what any of these columns mean. The information provided with the dataset is as follows:

column A: past 28 week slope value

column B: past 48 week slope value

column C & D: number of positive returns divided by negative returns and expressed as a percentage for the last 28 weeks and 48 weeks respectively

column E: returns this week

Even with the provided info I am struggling to understand what it means since I have no idea regarding stock market predictions and even after searching for a long time I cannot find anything similar to this dataset. It would be a great help if someone can point me in the right direction since I cannot make any predictions when I'm clueless about the dataset!!!

Please don't remove this question , I'm really desperate!

• This is for a class in what kind of forecasting method(s)? Regression, machine learning, .... A regression of column E on the preceding 4 columns might be a start. – noob2 Nov 5 '20 at 13:42
• I tried using linear regression and random forest to predict column E but ended up with literally zero accuracy and I'm trying to use the ARIMA time series approach. However, I first really want to understand what the column names mean! – user3918807 Nov 5 '20 at 14:42
• (Guessing, just from the name). 28wk slope: the slope of a regression line fitted to the last 28 weeks of stock prices (positive: stock has trended up last 28 weeks, negative: stock has been going down). 48wk slope: similar, but over 48 weeks (approx 1 year). WinLoss28: the number of up days in the last 28 weeks, in percentage, a number between 0 and 100). WinLoss48: similar, but over 48 weeks. – noob2 Nov 5 '20 at 15:49
• So called "momentum traders" like to buy stocks that have gone up in last 6 or 12 months and avoid those which have gone down. (A technique popularized by Jegadeesh and Titman). – noob2 Nov 5 '20 at 15:52
• Predicting stock price movements is an incredibly difficult task. No surprise if you have trouble. In my experience ARIMA does not work well for stocks. Good luck. – noob2 Nov 5 '20 at 15:54