ben-rounds / GA_portfolio

A portfolio of data science projects for the General Assembly Data Science course
0 stars 0 forks source link

HW1 Feedback #1

Open lemonsoup opened 8 years ago

lemonsoup commented 8 years ago

Status: Pass (Homework is graded on "Pass" or "Needs Improvement") Comments: Great work Ben! Your analysis is thorough and clear, with visuals that support your conclusions effectively. Clear labeling on your plots communicate your points very well and add to your analysis. Thumbs up to taking it a step further with SKlearn! I encourage you to check out the logistic regression, or logit regression -- linear regression is not great when the outcomes are categorical (like 0/1, male/female) it's better used on continuous data. (Spoiler alert for future classes!)

masongallo commented 8 years ago

Echoing Lema's comments above - loving the initiative! We will see in the logistic regression class why running linear regression in this case prob won't get us what we want. Your analysis for what predicts survival is spot on. Your idea for grouped data by individuals will come into play when we cover tree-based methods - we can use a series of logical statements about an individual to predict whether they survive. I encourage you to check out what some of your classmates did with the Names feature, along with thoughts around 'family' using Sibsp and Parch.

I would also encourage you to use Markdown for your written analysis rather than comments when you finalize your work for your portfolio. Great job!