BIZNIQUE / LINEAR ALGEBRA / 02
Represent data as vectors and matrices
Represent data as vectors and matrices
Decide what a row, a column and a missing value mean before choosing an algorithm.
Embedded Files
THE IDEA
Read the example.
Read the example.
In this example, each row is one observation and each column is one numerical feature. Multiplying by a weight vector produces one score per row. Changing a weight affects every row through that feature.
Xw = predictions
Use it well
Use it well
A category code is not automatically a quantity. Keep missing values distinct from measured zeros, and preserve the feature order used during training.
TRY IT YOURSELF
A quick check.
A quick check.
For the row (1, 2, 3) and weights (1, 0, −1), compute the score.
Check your reasoning
Check your reasoning
The score is −2: 1×1 + 2×0 + 3×(−1).
CONTINUE EXPLORING
Explore this topic.
Explore this topic.
Page updated
Google Sites
Report abuse