Deep Learning CNN In Python
CNN in Python to classify special images on the Dataset.
The methods used in the code are those described in details we provide after the agreement and may be improved.
The Training accuracy must be better than 95% on any label
The validation accuracy must be better than 85% on any label
The testing accuracy must be better than 80% on any label
The code should be heavily commented.
Each Line comment should have at least two pieces of information from the following:
1) what is the line of code doing
2) why line of code is doing that.
3) What is the aim to be achieved later from this line of code
The code must be in functions. NO LOOSE CODE AT ALL.
Functions must do one simple task. Except when it is nessary or convenient to do more than one task
All functions must be parameterized.
Global variables must be used as parameters to functions calls
Global variables must NOT be used in code lines directly.
Functions must be commented:
Function method or algorithm must be summerized/explained
Function parameters must be explained
The code should include training and prediction and confusion matrix and any other visualization.
The code should be in Python WITHOUT KERAS
All code to be delivered in less that 2 weeks from the start of this agreement. This is critical.
The code to be done on jupyter notebook shared with me on daily basis
It is essential I see the code every day, so that I can understand the code and ask questions.
I will review the code every day, ask questions about the code, and expect good informative replies.
This communication will done in a method as mutually agreed
The money is paid AFTER the delivery of the CNN and code is fully tested by me and meets all the requirements above
31 freelancer chào giá trung bình£565 cho công việc này
Hi, there. I am a machine learning engineer with 5+ years of hands-on experience in Python and CNN. Carefully checked your requirements. Hope to work for you with this project. Best regards, Sgadou