We are looking for a Data Scientist able to provide actionable insights from analyzing company data. The ideal candidate is experienced in machine learning for image recognition, particularly in Python 3. They must have strong experience in extracting, manipulating, and analyzing relevant data records within a database using SQL.
The task is to create a model to classify product images with their product identification number. A deliverable trained model and its code to train/test it is expected after results are presented. Further details regarding accessing the company database and communication with staff will be provided upon acceptance. The right candidate should be comfortable speaking in a non-technical manner to ask questions, raise concerns about the project, explain work, and/or present results.
Responsibilities:
- Mine company databases for labeled images.
- Train an image-recognition model to classify product images.
- Validate model quality and explain validation procedure.
- Communicate results through presentation.
- Deliver code used to generate model to company employees.
Qualifications:
- Has basic programming skills in Python 3.
- Familiar with SQL server enough to understand directions about data access.
- Can train an image classifier (TensorFlow, PyTorch, or Keras).
- Experience in troubleshooting and optimizing neural network architecture.
- Able to assess model performance through confusion matrix analysis (ROC, precision, recall, etc.).
- Comfortable presenting findings and describing procedures taken.
Skills & Expertise
- Deep Learning
- Image/Object Recognition
- Python
- SQL
About the recuiterMember since May 20, 2018 Pankaj Doot
from Ordu, Turkey