python programming for lotto result prediction

i need an expert in Python programming to write a deep learn program to extract 15 nos out of 49 (adjustable based on different lotto) nos with a confidence level of 95 percent. Result of lotto draw is 7 nos always. Means the 15 no predicted will contain the 7 no with a confidence of 95 percent. The program should be able to update itself from the last update in lotto results to the current outcome (based on the date and last lotto draw number). Based on the lotto results to predict using statistical tools and probability outcome to produce 15 most likely numbers to occur in the following draw with a confidence level of 95% The requirements for this project: 1. program must be able to automatically update itself on the last draw to the current draw no to the database in excel (outcome one). 2. Program to allow me to adjust the lotto nos (default 49 but can be adjusted to xx) 2. program must be able to produce with a confidence level of 95% 15 nos most likely to occur in the next draw of 7 nos. 3. Program must be able to machine learn to improve the outcome Milestones: 1. python program to be able to extract draw results from the internet (based on link provided) and download and output to excel file and based on this to predict 15 most likely nos to be drawn in the next lotto with confidence of 95% - 50% payment 2. backtest the program to confirm result as required above = 30 percent 3. output of program source code - 20 percent