Alternating Least Squares (ALS) adalah algoritma faktorisasi matriks yang sering digunakan dalam sistem rekomendasi, terutama untuk collaborative filtering hibrida. Algoritma ini memecah matriks rating pengguna-item menjadi produk dari dua matriks berperingkat lebih rendah, sehingga mencapai reduksi dimensi, mengisi nilai yang hilang, dan menemukan preferensi pengguna laten serta karakteristik item.
Sumber daya komputasi yang didukung
MaxCompute
Flink
Input dan output
Port input
Komponen upstream yang didukung:
Port output
Faktor Pengguna dan Item sesuai dengan Als Rating.
Konfigurasi komponen
Tambahkan komponen Als Matrix Factorization pada halaman pipeline dan konfigurasikan parameter berikut:
Kategori
Parameter
Deskripsi
Fields Setting
user column name
Nama kolom ID pengguna di tabel input. Data di kolom harus bertipe BIGINT.
item column name
Nama kolom item di tabel input. Data di kolom harus bertipe BIGINT.
rating column name
Nama kolom yang berisi skor yang diberikan oleh pengguna untuk item di tabel input. Data di kolom harus bertipe numerik.
Parameters Setting
num factors
Jumlah faktor. Nilai valid: (0,+∞). Nilai default: 10.
Number of iterations
Jumlah iterasi. Nilai valid: (0,+∞). Nilai default: 10.
Regularization coefficient
Koefisien regularisasi. Nilai valid: (0,+∞). Nilai default: 0,1.
check box
Menentukan apakah akan menggunakan model preferensi implisit.
alpha parameter
Koefisien preferensi implisit. Nilai valid: (0,+∞). Nilai default: 40.
Output table lifecycle
Siklus hidup tabel model output. Unit: hari.
Tuning
Number of Workers
Jumlah node pekerja. Nilai valid: 1 hingga 9999.
Node Memory, MB
Ukuran memori setiap node pekerja. Nilai valid: 1024 hingga 65536. Unit: MB.
Contoh
Jika Anda menggunakan data berikut sebagai input untuk komponen Als Matrix Factorization, Anda dapat memperoleh faktor pengguna dan faktor item berikut.
Data Input
user_id
item_id
rating
10944750
13451
0
10944751
13452
1
10944752
13453
2
10944753
13454
2
10944754
13455
4
... ...
... ...
... ...
Tabel Faktor Pengguna Output
user_id
faktor
8528750
[0.026986524,0.03350178,0.03532385,0.019542359,0.020429865,0.02046867,0.022253247,0.027391396,0.018985065,0.04889483]
282500
[0.116156064,0.07193632,0.090851225,0.017075706,0.025412979,0.047022138,0.12534861,0.05869226,0.11170533,0.1640192]
4895250
[0.038429666,0.061858658,0.04236993,0.055866677,0.031814687,0.0417443,0.012085311,0.0379342,0.10767074,0.028392972]
... ...
... ...
Tabel Faktor Item Output
item_id
faktor
24601
[0.0063337763,0.026349949,0.0064828005,0.01734504,0.022049638,0.0059205987,0.008568814,0.0015981696,0.0,0.013601779]
26699
[0.0027524426,0.0043066847,0.0031336215,0.00269448,0.0022347474,0.0020477585,0.0027995422,0.0025390312,0.0033011117,0.003957773]
20751
[0.03902271,0.050952066,0.032981463,0.03862796,0.048720762,0.027976315,0.02721664,0.018149626,0.0149896275,0.026251089]
... ...
... ...