We from taxtastic feel very honored to be mentioned in the annual π½ππππππ βππ€πππ ππ₯ππ£π€ report from π½ππππππ ππππ π§ππ₯π π£π€.
FinTech nerds can pre-order a version with: https://t.co/KM2cOyVXko
Thanks for selecting us @FinTech Innovators!
βπ πππ€π₯ππππ€, ππ¦π€π₯ πππ‘π‘πͺ πΈπ ππππππππ₯π€!
Most people post about their successes on LinkedIn, but aren't mistakes and failures the actual turns we learn most? I decided to share my machine learning failures from now on when one arises,starting with today
First, create a validation dataset and balance this one with the data you have. Later upsample the training dataset! Doing this in the wrong order will lead to some of your data points ending up in both your training and validation datasets.
@CryptoSchutt Great Question! I think a lot of text problems are solved with modern language models where a simple old school regular expression would suffice!
I hope that helps! ;)
π€AI Nerd Alert:
Using neural networks does not make you an AI startup! Andrew Ng: "During the rise of the internet, many companies made the mistake of believing that tacking a website onto the side of their business made them an internet company.
Conclusion:
AI is a hammer, but not every problem is a nail! Use Deep Neural Networks where it is appropriate to use them! Organizing your company around solving problems where AI is great at solving these problems makes you an AI startup.