Scribd
Upload
81 views1 page

Rubric For Assignment Questions-Deep Learning

The document outlines a rubric for evaluating assignment questions based on seven criteria: understanding of the problem statement, model design and implementation, use of techniques/tools/optimization, data handling and preprocessing, training, evaluation and metrics, result analysis and explanation, and code/documentation. Each criterion is graded on a scale from excellent to needs improvement, with specific descriptions for each level. The rubric provides a clear framework for assessing the quality of student submissions in a structured manner.

Uploaded by

rsd
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
81 views1 page

Rubric For Assignment Questions-Deep Learning

The document outlines a rubric for evaluating assignment questions based on seven criteria: understanding of the problem statement, model design and implementation, use of techniques/tools/optimization, data handling and preprocessing, training, evaluation and metrics, result analysis and explanation, and code/documentation. Each criterion is graded on a scale from excellent to needs improvement, with specific descriptions for each level. The rubric provides a clear framework for assessing the quality of student submissions in a structured manner.

Uploaded by

rsd
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

RUBRIC FOR ASSIGNMENT QUESTIONS

Criteria Excellent (Full Marks) Good (Partial Needs Improvement Marks


Marks) (Low or No Marks)
1. Understanding Demonstrates clear Shows basic Misunderstands or 2
of Problem understanding of the understanding incorrectly interprets
Statement problem; translates it but has minor the problem.
(PO1, PO12) accurately into model conceptual
or code logic. errors.
2. Model Design Correctly applies neural Mostly correct Incomplete, wrong 4
& network design with architecture or fails to
Implementation (DNN/CNN/RNN) with minor flaws in implement key
appropriate layers, structure or logic. components.
activations, and
structure.
3. Use of Proper use of Some techniques No advanced 3
Techniques / techniques like used correctly, techniques applied or
Tools / dropout, regularization, others partially or applied incorrectly.
Optimization optimizer comparison, incorrectly.
hyperparameter
tuning, etc.
4. Data Handling Preprocesses data Basic Data is not properly 2
& Preprocessing efficiently (scaling, preprocessing preprocessed or
encoding, applied with contains errors.
augmentation, etc.), minor issues.
handles input/output
well.
5. Training, Trains the model Evaluation done Missing or incorrect 3
Evaluation & correctly, evaluates but lacks proper training/evaluation
Metrics with proper metrics visualization or process.
(accuracy, confusion interpretation.
matrix, loss curves,
etc.) and presents
results clearly.
6. Result Analysis Thorough analysis of Some Little or no analysis of 3
& Explanation results, model comparison and results or incorrect
performance explanation interpretation.
comparison, and provided, lacks
reasons clearly stated. clarity or depth.
7. Code & Code is clean and well- Minor issues in Poorly written report 3
Documentation commented. Report is formatting or and/or poorly
well-structured with comments; documented code.
diagrams, explanation, report lacks some
and proper formatting. clarity.

You might also like