Learn how you can save hours of time on your machine learning and data science projects. 👉🏻 Link to document docs.datalumina.io/bJhyeJ4lHG... If you find this video helpful, consider subscribing @daveebbelaar
and not just ask, but know when the results/answer actually make sense as I've gotten incorrect answers from chat gpt before when I asked it to perform some stats for me.
Nice job and great video bro! Continue like this, appreciate it! However, I think there is an error in your code which significantly affects your model performance and accuracy making it of 85%, since you are supposed to split your dataset into training and testing sets before applying any transformations including the 'get_dummies()' method for handling categorical variables, because otherwise you cause a data leakage - where information from the test set inadvertently influences the training process, leading to overly optimistic performance estimates. And indeed, if you will restructure the process of your code and make the changes I described above, you will probably notice a significant lower accuracy of your model, reflecting real world cases.
This has changed the game entirely. From now on, schools should teach students how to use this AI and how to maximise desired outcomes of any known and unknown achievements. I will never be able to program anything without this AI again 😅
Can you make a video about how to be a freelance data scientis? How did you get started? Recommendations, how to get clients and how to actually concrete a business and general stuff like that. It'd be amazing
Hi Kunal, AI is definitely going to have an impact on the way we work, for all industries. But I like the example of accountants, who theoretically could have been replaced by software automation many years ago. However, we still need accountants for their expert judgement and specific knowledge to solve problems. I think the same will happen to Data Analytics/Science. It will be a tool for us to work faster and take on more projects.
hi dave. please share the tutorial about using python with visual studio code like this video, plus python environment setting and interface result. I wanna create machine learning codes using VSC too🙏🏻🙏🏻 thanks
bit weird, I was following along with this tutorial and when evaluating the best model to go with I got these results - Model: Ridge Regression Mean Squared Error: 51.64 R-squared: 0.80 ------------------------------- Model: Random Forest Mean Squared Error: 54.34 R-squared: 0.79 ------------------------------- Model: Gradient Boosting Mean Squared Error: 51.61 R-squared: 0.80 ------------------------------- Model: Support Vector Machine Mean Squared Error: 59.99 R-squared: 0.77 ------------------------------- Model: K-Nearest Neighbors Mean Squared Error: 68.27 R-squared: 0.74 ------------------------------- Gradient Boosting is the slight favourite, next I followed along with the tuning Hyperparameters with a grid search etc exactly as written in the video and I ended up getting a worse accuracy? Best Model: GradientBoostingRegressor(learning_rate=0.01, n_estimators=50, random_state=42) Best Mean Squared Error: 148.50 Best R-squared: 0.38 What's happened there?!
Ciao, sono Alessandro un personal Trainer e vorrei che tu mi aiutassi a sviluppare un GPTs che riesca a programmare in modo ottimale gli allenamenti in modo personalizzato per i miei clienti, basandosi solo sul mio materiale di studio. Fornendo le giuste progressioni, voglio che in base a quello che gli dico (anamnesi del soggetto) lui riesca a sviluppare un ottimo programma come faccio io così da aiutarmi ad essere più efficiente nel mio lavoro.
my python sucks balls; however I do have a masters degree in data sci (good grade too); do people think its cheating or will look stupid I I start issuing chat gpt to do most of my code at work; and people can see? I mean I know the theory and what to ask and edit code when required??? thoughts!!
Not at all! Use the tool to help you out and learn. As you work with it, you'll get a better understanding of Python. You can just ask it to explain the code as well. In my opinion, the best way to learn python for data science right now is to do projects slightly out of your comfort zone with the help of ChatGPT and GitHub Copilot.
As a data scientist There are way too many errors in approach here, when you work on real world data rather than showing a tutorial on curated data 1. No EDA, you just dove right in to a model Initial eda is crucial to understand what is wrong with the data and how it behaves 2. Leaky Fit-transform of dummy variables, and very little preprocessing overall 3. Tuning the best un-tuned model is not necessarily the best model overall Many models can be very bad untuned and drastically better when properly tuned Chat gpt is a powerful tool that can drastically improve the workflow of any industry But, as shown here, you always need to ask the right questions and keep checking your flow in order to get the best results