Perbandingan Model | Analisis Sentimen MBG

Filter Batch Dataset

🥧 Distribusi Data — Data MBG

5827

Total

1055

Positif

3302

Negatif

1470

Netral

📊 Grafik Perbandingan Metrik Evaluasi

🌲 Random Forest (Training Terakhir)

Akurasi	73.78%
Precision	74.52%
Recall	73.78%
F1-Score	74.05%
Data Train / Test	7968 / 1171
Waktu Training	1216.2 detik

📍 KNN (Training Terakhir)

Akurasi	52.78%
Precision	64.72%
Recall	52.78%
F1-Score	55.15%
Data Train / Test	7968 / 1171
Waktu Training	70.136 detik

Confusion Matrix — Random Forest

Aktual \ Prediksi	positif	negatif	netral
positif	150	39	23
negatif	46	527	91
netral	48	60	187

Confusion Matrix — KNN

Aktual \ Prediksi	positif	negatif	netral
positif	148	17	47
negatif	139	349	176
netral	141	33	121

Classification Report — Random Forest

Label	Precision	Recall	F1-Score	Support
positif	61.48%	70.75%	65.79%	212
negatif	84.19%	79.37%	81.71%	664
netral	62.13%	63.39%	62.75%	295

* Support = jumlah data uji per kelas · F1-Score = harmonic mean precision & recall

Classification Report — KNN

Label	Precision	Recall	F1-Score	Support
positif	34.58%	69.81%	46.25%	212
negatif	87.47%	52.56%	65.66%	664
netral	35.17%	41.02%	37.87%	295

* Support = jumlah data uji per kelas · F1-Score = harmonic mean precision & recall

Kesimpulan Perbandingan

Berdasarkan hasil training terakhir pada dataset batch "Data MBG", algoritma Random Forest menghasilkan akurasi lebih tinggi yaitu 73.78%, unggul 21 poin persen dibandingkan algoritma pembanding.

Seluruh Riwayat Training

🌲Random Forest

Waktu	Batch	Algoritma	Akurasi	Precision	Recall	F1
13/09/2026 21:35	Data MBG	🌲 RF	73.78%	74.52%	73.78%	74.05%

📍K-Nearest Neighbor (KNN)

Waktu	Batch	Algoritma	Akurasi	Precision	Recall	F1
13/09/2026 21:38	Data MBG	📍 KNN	52.78%	64.72%	52.78%	55.15%