Automated AI tool including: recommender system, deep Learnings ...
Project description
witcher
Witcher is an automated, driverless tool to build an error-free machine learning application.
Witcher is an automated AI application designed to speed up the data processing phases. The current witcher has :
- Recommender: will provide you with a comprehensive product recommendation using product popularity, product similarity and user similarity models.
- StockMarket : Will read the stock information from Yahoo API and analysis the Stock prices using ARIMA Time series
- FileChooser : Automated FileChooser function, Filechooser will read your data regardless it format and will provides you a Datafram. Currently witcher can read csv,xls,xlsx,sas,and images automaticaly
- ImageToCartoon : A fun project will read your images and do some processing such as image blurred, edges extraction, image to cartoon,...
Magical data processing, predictive, and deep learning models will be joining witchers functions very soon :)
I hope you will enjoy using the witcher library
Thank you.
Babak.EA Founder and CEO: AI Forest Inc
How to run :
using Jupyter notbook
- import wirtcher
- from witcher import Recommender
- from witcher import StockMarket
- from witcher import FileChooser
- from witcher import ImageToCartoon
from witcher import FileChooser
FileChooser.Filechooser ==>
will read the file and will returen the filename, filepath, and dataframe for
CSV, XLS,XLSX, SAS, and Images
Reprort=witcher.FileChooser.filechoose()
Report
report.files == > file path
report.df ==> dataframe
from witcher import FileChooser, ImageToCartoon
img=FileChooser.Filechooser()
img
image=img.df or
image=ImageToCartoon.Img_Reader(img.files[0]) # read image and return numpy vector
ImageToCartoon.ImageShow(image) ### Show the image
Image_D=IMG_D=ImageToCartoon.Decolorization(image)
Blurred=Blurred(image)
edges=ImageToCartoon.Edgedetection(image)
Bluured=ImageToCartoon.Blurred(image)
Cartoon=Cartoon(bluured=Blurred,mask=edges)
or
Cartoon=Cartoon(bluured=Blurred,decolor=Image_D,mask=edges)
from witcher import StockMarket
select your stock and starting date to end date
df=StockMarket.Stock_Reader(stock=["AC.TO"],Period="1D",Start_date="2010-01-01",End_date="Today")
df=StockMarket.Dataset_Spliter(df,col="Close",split=.1,Forecasting=True)
select the column you want to analysis and pass it to the finction
from witcher import Recommender
submit your user products and run the recommender system
Model_generator=Recommender.Model_Generator(User Product dataframe,User Satisfaction DataFrame,User Informatiin dataframe ,knn_neighbors=number of the neighbors)
*** Witcher Recommender system work based on 3 different recommender systems, SVD, PCA, and KNN with a total accuracy of 80-85%
df=StockMarket.Dataset_Spliter(df,col="Close",split=.1,Forecasting=True)
Get Recommendation :
products=Recommender.Product_Recommender("User ID")
Recommender.pprint.pprint(products.report)
Recommendation for a new user:
products=Recommender.Product_Recommender("User Name",NEW_USER="True",USER_feature=[list of the selected producte by user]+[User incom]+[User satisfaction level] ( new user is 0))
Recommender.pprint.pprint(products.report)
How To Install
pip install witcher
source code : github
https://github.com/BabakEA/witcher
YouTube:https://www.youtube.com/channel/UCBqqRv8vWV3NZFF2tQV4e-w
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file witcher-0.0.41.tar.gz
.
File metadata
- Download URL: witcher-0.0.41.tar.gz
- Upload date:
- Size: 43.0 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.0 CPython/3.9.12
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 785e969c58e38f3ac04f1c44205092924981360084c6a75cfe52305da467da5b |
|
MD5 | ba616f0a4b8287c201308b7d7583e237 |
|
BLAKE2b-256 | fa99920c684c8168ef1f55744166ac46534dad17fd83595f4b6f4e2e33d260c2 |
File details
Details for the file witcher-0.0.41-py3-none-any.whl
.
File metadata
- Download URL: witcher-0.0.41-py3-none-any.whl
- Upload date:
- Size: 43.2 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.0 CPython/3.9.12
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | d15a52c757f27e5f65995a2bc7ac9d092502677e42ee050640686c8ea30d6fd0 |
|
MD5 | 7d8778d23d5a4dfc2a60d7381f01439d |
|
BLAKE2b-256 | 0899ad1a3373d6eeb860d893b3e682a471afdc8ba47beb2af87d643b9f9eb386 |