Add Mean Absolute Error Loss (#587)

Author: sudo-rsingh

src/machine_learning/loss_function/mae_loss.rs

@@ -0,0 +1,36 @@
+//! # Mean Absolute Error Loss Function
+//!
+//! The `mae_loss` function calculates the Mean Absolute Error loss, which is a
+//! robust loss function used in machine learning.
+//!
+//! ## Formula
+//!
+//! For a pair of actual and predicted values, represented as vectors `actual`
+//! and `predicted`, the Mean Absolute  loss is calculated as:
+//!
+//! - loss = `(actual - predicted) / n_elements`.
+//!
+//! It returns the average loss by dividing the `total_loss` by total no. of
+//! elements.
+//!
+pub fn mae_loss(predicted: &Vec<f64>, actual: &[f64]) -> f64 {
+    let mut total_loss: f64 = 0.0;
+    for (p, a) in predicted.iter().zip(actual.iter()) {
+        let diff: f64 = p - a;
+        let absolute_diff: f64 = diff.abs();
+        total_loss += absolute_diff;
+    }
+    total_loss / (predicted.len() as f64)
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn test_mae_loss() {
+        let predicted_values: Vec<f64> = vec![1.0, 2.0, 3.0, 4.0];
+        let actual_values: Vec<f64> = vec![1.0, 3.0, 3.5, 4.5];
+        assert_eq!(mae_loss(&predicted_values, &actual_values), 0.5);
+    }
+}

src/machine_learning/loss_function/mod.rs

@@ -1,3 +1,5 @@
+mod mae_loss;
 mod mse_loss;
 
+pub use self::mae_loss::mae_loss;
 pub use self::mse_loss::mse_loss;

src/machine_learning/mod.rs

@@ -3,5 +3,8 @@ mod loss_function;
 mod optimization;
 
 pub use self::linear_regression::linear_regression;
+
+pub use self::loss_function::mae_loss;
 pub use self::loss_function::mse_loss;
+
 pub use self::optimization::gradient_descent;