ML Bits @ml_bits

Daily machine learning tips, tricks, and best practices in your feed by @burkov Joined November 2020

Tweets

45
Followers

591
Following

1
Likes

326

ML Bits @ml_bits

4 years ago

ML Bit #37 When communicating the performance numbers for your machine learning models to the stakeholders, do say that 3% error doesn't mean that the model will take the same input 100 times and fail 3 times. If all the 100 inputs are hard for the model, it will fail 100 times.

0 1 9 0 4

ML Bits @ml_bits

4 years ago

ML Bit #36 In machine learning, if there's an implementation error, and your model is suboptimal because of it, there's no way you can know that. Unless you implement one or multiple benchmarks one of which would beat your solution.

0 0 5 0 0

ML Bits @ml_bits

4 years ago

ML bit #35

BURKOV @burkov

4 years ago

ML bit #35

2 26 142 0 17

0 0 4 0 0

BURKOV @burkov

4 years ago

Several sponsorship spots are left in October in my weekly AI newsletter with more than 470,000 subscribers. If your company's business is based on AI or data or you organize an event in this domain, it may be a fit. If you are interested, please contact [email protected].

0 2 5 0 0

ML Bits @ml_bits

4 years ago

ML Bit #34 If your model must predict perfectly all examples from the training, validation, and test data, but you still need holdout data for model validation, then in production use a database lookup for all examples that you already have, and only use the model for new data.

0 1 9 0 0

BURKOV @burkov

4 years ago

Currently, three counterfeit sellers sell my book on @amazon : Blue planet UK, Prowisdombooks, and assis books. They took over the "Buy Now" and "Add to Cart" buttons and I lost almost all sales. Amazon's been aware of the situation for more than 10 days now. Beware!

0 4 13 0 0

Download Image

ML Bits @ml_bits

4 years ago

ML Bit #33 When you deliver a model, make sure that all stakeholders understand that your model's performance is guaranteed *only* when the model's input in production obtained following *exactly* the same procedure. Otherwise, there's only hope, no guarantee.

0 1 13 0 0

ML Bits @ml_bits

5 years ago

ML Bit #32 When choosing machine learning algorithms/models for your problem, choose those that are the fastest on your dataset with the default values of hyperparameters. This will allow you to test more hyperparameter configurations and will give you a strong baseline.

0 3 15 0 1

ML Bits @ml_bits

5 years ago

ML Bit #31 Most scikit-learn supervised learning algorithms for classification are prohibitively slow for a typical 2020's dataset. Except for SGDClassifier: super fast and accurate.

0 0 15 0 3

ML Bits @ml_bits

5 years ago

ML Bit #30 Be careful when buying data for machine learning: you can purchase crap generated by a poorly trained model disguised as an original dataset.

0 0 4 0 0

ML Bits @ml_bits

5 years ago

ML Bit #29 Model ensembling can be very powerful, especially when you combine orthogonal models: those trained on different types or sources of data or using different ML algorithms: different architectures of neural networks, tree-based algorithms, kNN, and SVM.

0 1 19 0 2

ML Bits @ml_bits

5 years ago

ML Bit #28 Word-based or character-based encoding for strings? Byte pair.

0 2 11 0 0

ML Bits @ml_bits

5 years ago

ML Bit #27 Use your model to find the wrong labels in your training data. If your model predicts a different label, verify that the label in the training data is correct, fix it if needed, and retrain the model.

0 2 9 0 0

ML Bits @ml_bits

5 years ago

ML Bit #26 When you need to compute the dot product between two vectors, never use a for-loop. Use numpy.dot instead. It will be ~100 times faster.

0 1 21 0 0

ML Bits @ml_bits

5 years ago

ML Bit #25 Use numba, pypy, or Cython to speed up your Python code.

1 3 26 0 0

ML Bits @ml_bits

5 years ago

ML Bit #24 Save code snippets that you frequently use to Github gists.

0 1 12 0 0

ML Bits @ml_bits

5 years ago

ML Bit #23 Classification can be 1) binary when the model outputs one label and that label can have one of two possible values, 2) multiclass when the output label is still unique but can be one of the N values, and 3) multilabel when the model can output several labels.

0 1 12 0 1

ML Bits @ml_bits

5 years ago

ML Bit #22 kNN is accurate and it's not slow.

1 0 10 0 1

ML Bits @ml_bits

5 years ago

ML Bit #21 Before training a neural network, train a baseline. XGBoost has a great performance out of the box and would serve as a good baseline. Other choices: Linear SVM, logistic regression, and kNN with k=3 or 5 and Euclidean distance.

0 0 15 0 1

ML Bits @ml_bits

5 years ago

ML Bit #20 If your neural network has two outputs or you would like to optimize two cost functions, it's possible by introducing another hyperparameter cost = gamma \times cost_1 + (1-gamma) \times cost_2. The value of gamma rages in (0, 1) and to be tuned on validation data.