Workshop Name: Software Forensics Through Application of Logging in Machine Learning Code
Description
Identify logging locations in a machine learning (ML) code.
Targeted Courses
Software Quality Assurance
Activities
Pre-lab Content Dissemination
We first need to know what to log and how to log. Let us use the following heuristics.
Heuristics on what to log
Resources
Exceptions
Change issues
Heuristics on how to log
Include the name of the identity provider or security realm that vouched for the username, if that information is available. Include the affected system component or other object (such as a user account, data resource, or file).
Include the status that says if the object succeeded or failed.
Include the application context, such as the initiator and target systems, applications, or components.
Include “from where†for messages related to network connectivity or distributed application operation.
Include the time stamp and time zone help answer “when.†The time zone is essential for distributed applications.
Heuristics on how not log
Do not log user names and passwords
Do not log sensitive information, such as credit card numbers or social security numbers Do not log nullable objects
In-class Hands-on Experience
Open https://docs.python.org/3/library/logging.html
Learn about the common logging methods with Python
Integrate logging for simple.py. Target the basic important things to log.
import
Code will be saved and uploaded in this repo.
Demo will be recorded and uploaded on CANVAS.
Assignment 3 (Post Lab Experience)
1. poisoning attacks that are introduced through passing in erroneous dataset files into ML code
2. model tricking that are results of models that are being attacked, which gives erroneous results. Incorrect prediction performance is an indicator of ML models being attacked.
3. Assist Dolly by writing logging code in the correct locations in workshop9.py with comments so that the two above-mentioned issues are logged for all provided ML functions. After writing the code, put in comments to justify your code, save workshop9.py.
4. You have the freedom to choose the locations that you think needs to be logged
Sample Input/Output:
Input:
workshop9.py
Expected Output: Code and Comments
iris = datasets.load_iris() // initiated logging logging.basicConfig(filename='app.log', filemode='w', format='%(name)s - %(levelname)s - %(message)s')
