Linear Algebra

机器学习常用线性代数概念

Linear Algebra

线性代数为机器学习模型提供了统一的表示语言。

Matrix multiplication

对于 $sequence( [A], [ ], [∈], [ ], attach(base: [ℝ], t: sequence([m], [ ], [×], [ ], [n])), )$ 和 $sequence( [B], [ ], [∈], [ ], attach(base: [ℝ], t: sequence([n], [ ], [×], [ ], [p])), )$ ：

$sequence( attach( base: lr(body: sequence([(], [A], [ ], [B], [)])), b: sequence([i], [ ], [j]), ), [ ], [=], [ ], attach(base: [∑], t: [n], b: sequence([k], [=], [1])), [ ], attach(base: [A], b: sequence([i], [ ], [k])), [ ], attach(base: [B], b: sequence([k], [ ], [j])), )$

矩阵乘法在注意力机制中用于计算相关性。

Linear Algebra

Matrix multiplication

Related notes