Review 1 - Week 1 ML Project

[okothchristopher]

General Comments

• Structure your notebook appropriately, with folders to each project, so you can have within the folder a Readme file, the project notebook file and the data
• Use this as an example
• For this you can use GitHub desktop.
• Leave out these code blocks, delete them entirely, when submitting projects pip install -U scikit-learn
• Structure your notebook as we have done in class, with sections and you can even number them for ease of navigation.

Project Specific Comments

• When getting a peak at your data, use df.head(). Otherwise your notebook becomes clanky with outputs for each cell.
• You have not indicated what strategy you are using for the missing values.
• This Assumption Moore's Law still holds, especially in GPUs. is wrong, I think this is for CPUs only. And be sure to explain your understanding of Moore's law.
• Use the website, provided in the readme file and replicate those graphs.
• This assumption Dannard Scaling is still valid in general. is okay, however, you need to treat your reader as someone who has never heard of it and explain to them, what doe the term even mean and why are you using TDP and Die Size to validate it.
• Some of these graphs, for example for this Process Size for Intel, AMD and Nvidia lies in comparatively lower range than for ATI and other vemdors are inordinately big. Resize appropriately.
• Explorations like these _semiconductors['Type'].valuecounts() should note be left in the notebook.
• Overall, good job on the project, but you need to work on the comments above to make it better.

[Doclikam]

Thank you for your feedback. I will see to it.

On Wed, 8 May 2024 at 15:42, okoth christopher @.***> wrote:

General Comments

• Structure your notebook appropriately, with folders to each project, so you can have within the folder a Readme file, the project notebook file and the data
• Use this as an example image.png (view on web) https://github.com/Doclikam/machine-learning-I/assets/52817534/1f43a2ea-73c8-4ba6-b8f8-5cb206a19982
• For this you can use GitHub desktop.
• Leave out these code blocks, delete them entirely, when submitting projects pip install -U scikit-learn
• Structure your notebook as we have done in class, with sections and you can even number them for ease of navigation.

Project Specific Comments

• When getting a peak at your data, use df.head(). Otherwise your notebook becomes clanky with outputs for each cell.
• You have not indicated what strategy you are using for the missing values.
• This Assumption Moore's Law still holds, especially in GPUs. is wrong, I think this is for CPUs only. And be sure to explain your understanding of Moore's law.
• Use the website, provided in the readme file and replicate those graphs.
• This assumption Dannard Scaling is still valid in general. is okay, however, you need to treat your reader as someone who has never heard of it and explain to them, what doe the term even mean and why are you using TDP and Die Size to validate it.
• Some of these graphs, for example for this Process Size for Intel, AMD and Nvidia lies in comparatively lower range than for ATI and other vemdors are inordinately big. Resize appropriately.
• Explorations like these semiconductors['Type'].value_counts() should note be left in the notebook.
• Overall, good job on the project, but you need to work on the comments above to make it better.

— Reply to this email directly, view it on GitHub https://github.com/Doclikam/machine-learning-I/issues/1, or unsubscribe https://github.com/notifications/unsubscribe-auth/BGBI5SAC3U63JMYFWBTXXN3ZBIMTPAVCNFSM6AAAAABHM4VK7SVHI2DSMVQWIX3LMV43ASLTON2WKOZSGI4DKNJQGYYTSMQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>