/machinelearningchallenge

Open Challenge for hiring at drivebuddyAI for AI/ML/DS positions.

Welcome to drivebuddyAI Programming Challenge!

Your Goal is to predict the genuiness of image of Pavbhaji(an Indian food item) based on the metadata provided in json file.

GIVEN :

Instagram posts data for #pavbhaji

----Dir --- dataset |----- images (Segregated images in "0" & "1") |----- Json file

Where json file consist of entries for post and its metadata.

TASK:

  • Use the Segregated data into positive(pavbhaji)("/dataset/1") and negative (non-pavbhaji)("/dataset/0") dirs
  • Build a model that could Distinguish between pavbhaji and non-pavbhaji images from ("/dataset/pavbhaji") json entry
  • Attain a level of Accuracy

Deliverable

Your deliverables are:

  • Source Code (.py or .ipynb) with Accuracy and Performance Parameters
  • Explain why you choose particular Model
  • What are the challenges you faced while working on the data?
  • How did you overcome those and what approach you try to solve the problem? And mail it us at connect(at)drivebuddyai(dot)com

Evaluation

  • We will evaluate you based on
    • Your Coding & Achieved level of accuracy
    • Your Problem Solving approach
    • Time Taken to Solve the Problem

Follow us on Twitter and Tag us @drivebuddyAI if you have completed the challenge and we'll call you for a coffee.