Skip to content

LETUS, used car price predication system with machine learning

Notifications You must be signed in to change notification settings

Georgesarkis/dataIntensiveAI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Please check DIT824-Team1-Report.pdf for informations related to the system.

Team01

Team Memebers

  1. George Sarkisian
  2. Martin Stanchev
  3. Majed Dalain
  4. Amjad Alshihabi
  5. Mohammad derham

Sprints

Sprint1

  • check available datasets and brainstorm to figure out what to use in our ML model. @George @Martin @Mohammad @Amjad @Majed
  • Add the Markdown file for first week to Gitlab. @George
  • Write code of conduct and documentation. @George
  • Present idea of the projcet the the teacher and TA's. @George @Martin @Mohammad @Amjad @Majed
  • start writing assig 1. @George @Martin @Mohammad @Amjad @Majed
  • Sprint retro. @George @Martin @Amjad @Majed
  • Sprint Review. @George @Martin @Amjad @Majed

Sprint2

  • Sprint Planing. @George @Martin @Amjad @Majed
  • Finish writing assignment 1 and submit it.@George @Martin @Amjad @Majed
  • Make script to clean up the data. @Martin
  • Check how to convert .csv file to SQLite. @Amjad
  • Do a research about different regression algorithm e.g. cart. @Majed
  • Send an email to the teacher and the TAs describing what the tools we are planning to use in our front end development. @George
  • Divide the dataset into 3 sets of data so it can later used by the admin to validate the ML model. @George
  • Study the data to find the best way to perform feature engineering @George
  • do a brainstorming session and discuss possible feature engineering @All

Sprint3

  • Decide on the DB schema. @All
  • Finish data validation schema and integrate with existing code. @Amjad
  • Convert all data to SQLite. @Amjad
  • Feature engineering (median) on HP based on car model, Mileage based on manufacture year. @George
  • Feature engineering on model and maker, on blocket dataset. @Martin
  • Decide on model evaluation criteria. Integrate it with existing code. @Majed
  • Feature engineering on categorical features - model and maker. @Majed
  • When the model is finished, fix the code. @Martin
  • Init implementation of frontend. @Majed

Sprint4

  • Rewrite the assig. 1 and resubmit. @ALL
  • Implement & intigrate server with falsk with simple APIs for GET & POST. @Majed @George
  • Check & implement a way to understand new input from the user to get prediction from the model(labelEncoder). @Martin
  • Read data from SQLite. @Amjed
  • Add new columns for the date of the data has been added to the database. @Amjed
  • Make a funcation to only retrive the data for a special date. @Amjed
  • Implement feature eng. on the data that user will enter to predict price. @George
  • Create the https request to POST the csv file to the backend. @George
  • Implement Model and Maker filed in frontend with autocomplete. @Majed
  • Clean MakeModel.json. @Martin

Sprint5

  • Create API endpoint to retrain the model with a specific date. @Majed POST request with body - from and to date
  • Save date as timestamp in the database. @Amjad
  • Create function to get all entries between two dates from the database. @Amjad
  • Implement Model and Maker filed in frontend with autocomplete. @Martin @Majed
  • Fix the bug in label encoder. @George
  • Fix the bug in predicting new values(Always the same price). @George
  • Fix the bug in the lambda median function, when the user doesn't input hp. @George
  • Add an admin page in the frontend. @Martin
  • Include hp in the user predict. @Martin
  • When doing the validation for the admin, draw a plot. @Amjad
  • Write code to retrain the model based on data from specific date from the database. @George

Sprint6

  • Unit testing. @George
  • Prepare clean jupyter notebook file for the model. @George
  • add versioning to the model. @Majed
  • autocomplete for car model. from previous sprint. @Majed
  • add visualization to Admin page. @Amjad
  • integrating the code.@Martin

Sprint7

  • Update the report. @George @Martin @Amjad @Majed
  • Submit the report.
  • add versioning to the model. @Majed
  • add visualization to Admin page. @Amjad

Attendencies§

Team Memebers 04/11/2019 07/11/2019 12/11/2019 14/11/2019 19/11/2019 20/11/2019 25/11/2019 26/11/2019 28/11/2019 02/12/2019 03/12/2019 05/12/2019 09/12/2019 10/12/2019
George Sarkisian X X X X X x x x x x x x x x
Martin Stanchev X X X X X x x x x x x x x x
Majed Dalain - X X X X x x x - x x x x x
Amjad Alshihabi X X X X X x x - x x x x x x
Mohammad derham X X - - - - - - - - - - - -

About

LETUS, used car price predication system with machine learning

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published