Predict Chart Growth [on hold]

Lets say I have 500 stock charts: They show the value of a share for a specific date:

With Machine Learning I would like to predict the value for a abitrary stock on the next Date.

I thought I could train my model by looking on the growth rates of the stock values. Lets say I have a stock that costs:

Day1: 100 Day2: 95 Day3: 110 Day4: 110 Day5: 105 Day6: 110  NextDate: My prediction

Then I would look on the growth rates:

Day1: 0% Day2: -5% Day3: 15.78% Day4: 0% Day5: -4.54% Day6: 4.76%  NextDate: My prediction

And would train my model by looking always on 3 following dates and using the forth as label. That would be:

X: [0, -5, 15.78]  Y: 0
X: [-5, 15.78, 0]  Y: -4.54
X: [15.78, 0, -4.54] Y: 4.76

Then to predict the value on the next date I would input:

next_date_growth = model.predict([0, -4.54, 4.76])

I am beginner in machine learning, how would you predict stock prices?And what Sklearn algortithm should I use? Thanks