87 points by mltutor 5 months ago flag hide 24 comments
original_poster 5 months ago next
@community_member, I'm impressed! I'll download the app and check it out. Great work!
critical_thinker 5 months ago prev next
I noticed that you used a train-validation split instead of a traditional train-test split. Can you explain why?
original_poster 5 months ago next
@critical_thinker, I used a train-validation split because I wanted to use the validation set to tune hyperparameters and evaluate the model's performance more accurately. I found that it helped me make better decisions about model selection and avoid overfitting. It's not always necessary to use a train-validation split, but it can be helpful in many cases.
john_doe 5 months ago prev next
Great post! I've been looking for a way to get into machine learning without a ton of prior experience. The portfolio looks like a great resource. Thanks for sharing!
jane_doe 5 months ago next
I'm glad you found the portfolio helpful! I'm excited to see the contributions you make. Stay in touch!
random_user 5 months ago prev next
Hey, just wanted to let you know I forked your repo and will be making some contributions soon. Keep up the good work!
original_poster 5 months ago next
@random_user, I appreciate it! Let me know if you need any help or guidance as you contribute.
user5 5 months ago prev next
I've been working through the portfolio, and I have a question about the linear regression section. Why did you choose that specific algorithm for this particular dataset?
original_poster 5 months ago next
@user5, good question! I chose linear regression for its simplicity and interpretability. I wanted to make sure that beginners could easily understand the model and its outputs. I also found that it worked quite well for this dataset.
another_user 5 months ago prev next
I'm having trouble setting up the environment. Are there specific versions of the dependencies that I need to install?
original_poster 5 months ago next
@another_user, I'm sorry to hear that! I recommend using conda to create an isolated environment for the project. Here are the dependencies and their versions: - Python 3.7 - NumPy 1.19.5 - Pandas 1.2.3 - Scikit-learn 0.24.2 - Matplotlib 3.3.4
third_party 5 months ago prev next
I created a web version of the portfolio using Streamlit. Check it out here: <https://personalized-ml-portfolio.streamlit.app/>
original_poster 5 months ago prev next
@third_party, wow, thank you for doing that! It looks great.
sixth_sense 5 months ago prev next
Have you considered using GitHub Actions or a similar tool for continuous integration and deployment?
original_poster 5 months ago next
@sixth_sense, I'm actually in the process of setting up GitHub Actions now. I'll add a note about it to the repo. Thanks for the suggestion!
new_user 5 months ago prev next
I'm new to machine learning, and I really appreciate you making this portfolio available for beginners. Is there a good place to learn more about the tools and techniques you used?
original_poster 5 months ago next
@new_user, I'm glad it's helpful! There are many great resources for learning machine learning, but I recommend starting with a few books like 'Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow' and 'Python Machine Learning: Machine Learning and Deep Learning with Python, scikit-learn, and TensorFlow 2.' Also, Kaggle has many datasets and tutorials that you can use to practice your skills.
another_forker 5 months ago prev next
I forked the repo and added a section on deep learning. Take a look and let me know what you think!
original_poster 5 months ago prev next
@another_forker, that's awesome! I'd love to take a look and incorporate any changes that make sense for the portfolio. Let's collaborate on a pull request.
community_member 5 months ago prev next
I built an Android app for the portfolio using React Native and Flask. It's available on the Play Store now. Check it out and let me know what you think!
collaborator 5 months ago prev next
I contributed a few changes to the repo, including a section on data cleaning and preprocessing. Let me know what you think!
original_poster 5 months ago next
@collaborator, I love the new section on data cleaning and preprocessing! It's a crucial step in the machine learning pipeline, and it's great to see it included in the portfolio. thanks for the contribution!
masters_degree 5 months ago prev next
I completed the portfolio and had a great time doing it. I learned a lot and appreciated the clear explanations. I'm considering pursuing a masters degree in machine learning. Any recommendations on good schools or programs?
original_poster 5 months ago next
@masters_degree, I'm glad you enjoyed the portfolio! Many schools offer masters degrees in machine learning, and it's essential to choose a program that fits your needs and goals. I recommend looking at programs that have a strong focus on practical applications and projects. Some of the top programs include Carnegie Mellon University, Stanford University, MIT, and UC Berkeley. Also, check out online resources like Coursera, edX, and Udacity for online courses and programs in machine learning.