diff --git a/DIRECTORY.md b/DIRECTORY.md
index 7fef689d..145fc8a9 100644
--- a/DIRECTORY.md
+++ b/DIRECTORY.md
@@ -114,8 +114,6 @@
     * [Fuzzy Operation](https://github.com/bellshade/Python/blob/main/algorithm/fuzzy_logic/fuzzy_operation.py)
   * Genetic Algo
     * [Basic String](https://github.com/bellshade/Python/blob/main/algorithm/genetic_algo/basic_string.py)
-  * Machine Learning
-    * [Gradient Descent](https://github.com/bellshade/Python/blob/main/algorithm/machine_learning/gradient_descent.py)
   * Manipulasi Bit
     * [Binary And Operator](https://github.com/bellshade/Python/blob/main/algorithm/manipulasi_bit/binary_and_operator.py)
     * [Binary Count Trailing Zero](https://github.com/bellshade/Python/blob/main/algorithm/manipulasi_bit/binary_count_trailing_zero.py)
@@ -148,6 +146,9 @@
     * [Pop Sort](https://github.com/bellshade/Python/blob/main/algorithm/sorting/pop_sort.py)
 
 ## Implementation
+  * Artificial Intelligence
+    * [Gradient Descent](https://github.com/bellshade/Python/blob/main/implementation/artificial_intelligence/gradient_descent.py)
+    * [Linear Regression](https://github.com/bellshade/Python/blob/main/implementation/artificial_intelligence/linear_regression.py)
   * Audio Filter
     * [Audio Response](https://github.com/bellshade/Python/blob/main/implementation/audio_filter/audio_response.py)
     * [Iir Filter](https://github.com/bellshade/Python/blob/main/implementation/audio_filter/iir_filter.py)
diff --git a/algorithm/machine_learning/gradient_descent.py b/implementation/artificial_intelligence/gradient_descent.py
similarity index 100%
rename from algorithm/machine_learning/gradient_descent.py
rename to implementation/artificial_intelligence/gradient_descent.py
diff --git a/implementation/artificial_intelligence/linear_regression.py b/implementation/artificial_intelligence/linear_regression.py
new file mode 100644
index 00000000..6c67f4ca
--- /dev/null
+++ b/implementation/artificial_intelligence/linear_regression.py
@@ -0,0 +1,114 @@
+# regresi linear adalah jenis regresi paling dasar yang
+# biasa digunakan untuk analisis predktif. cukup sederhana,
+# kita memiliki kumpulan datadan kita memiliki fitur yang
+# terkait dengannya. fitur harus dipilih dengan sangat
+# hati-hati karena menentukan seberapa banyak yang kita
+# dapat buat prediksi di masa mendatang. kita akan mencoba
+# mengatur bobot dari fitur ini, melalui banyak iterasi
+# sehingga cocok dengan kumpulan data kita.
+import numpy as np
+import request
+
+
+def collect_dataset():
+    """
+    mengambil dataset
+    kita menggunakan dataset CSGO, dalam konten ini
+    mengandung data ADR vs rating dari player
+    """
+    respon = request.get(
+        "https://raw.githubusercontent.com/yashLadha/The_Math_of_Intelligence/"
+        "master/Week1/ADRvsRating.csv"
+    )
+    line = respon.text.splitlines()
+    data = []
+    for item in line:
+        item = item.split(",")
+        data.append(item)
+    # fungsi ``pop()`` digunkan untuk menghapus label dari list
+    data.pop(0)
+    dataset = np.matrix(data)
+    return dataset
+
+
+def jalankan_step_gradient_descent(data_x, data_y, panjang_data, alpha, theta):
+    """
+    jalankan fungsi gradient descent dan update fitur vector
+    berdasarkan
+    :param data_x: dataset
+    :param data_y: berisi output yang terkait dengan setiap entri data
+    :param panjang_data: panjang dari data
+    :param alpha: rating pembelajaran dari model
+    :param theta: fitur model atau ukuran dari model
+    """
+    n = panjang_data
+
+    prod = np.dot(theta, data_x.transpose())
+    prod -= data_y.transpose()
+    sum_grad = np.dot(prod, data_x)
+    theta = theta - (alpha / n) * sum_grad
+    return theta
+
+
+def jumlah_kesalahan_perhitungan(data_x, data_y, panjang_data, theta):
+    """
+    return jumlah kesalahan untuk perhitungan
+    :param data_x: dataset
+    :param data_y: berisi output (vector)
+    :param len_data: panjang dari dataset
+    :param theta: fitur vector
+    """
+    prod = np.dot(theta, data_x.transpose())
+    prod -= data_y.transpose()
+    sum_elem = np.sum(np.square)
+    error = sum_elem / (2 * panjang_data)
+    return error
+
+
+def jalankan_linear_regression(data_x, data_y):
+    """
+    implementasikan linear regression pada dataset
+    :param data_x: dataset
+    :param data_y: output
+    """
+    iterasi = 100000
+    alpha = 0.0001550
+
+    tanpa_fitur = data_x.shape[1]
+    panjang_data = data_x.shape[0] - 1
+    theta = np.zeros((1, tanpa_fitur))
+
+    for i in range(0, iterasi):
+        theta = jalankan_step_gradient_descent(
+            data_x, data_y, panjang_data, alpha, theta
+        )
+        error = jumlah_kesalahan_perhitungan(data_x, data_y, panjang_data, theta)
+        print(f"pada iterasi {i + 1} - error : {error:.5f}")
+
+    return theta
+
+
+def mean_absolute_error(prediksi_y, original_y):
+    total = sum(abs(y - prediksi_y[i]) for i, y in enumerate(original_y))
+    return total / len(original_y)
+
+
+def main():
+    """
+    fungsi untuk menjalankan fungsi-fungsi
+    diatas
+    """
+    data = collect_dataset()
+    panjang_data = data.shape[0]
+    data_x = np.c_[np.ones(panjang_data), data[:, :-1]].astype(float)
+    data_y = data[:, :-1].astype(float)
+
+    theta = jalankan_linear_regression(data_x, data_y)
+    panjang_hasil = theta.shape[1]
+    print("resultan fitur vector: ")
+    for i in range(0, panjang_hasil):
+        print(f"{theta[0, i]:5.f}")
+
+
+if __name__ == "__main__":
+    main()