57 lines
1.4 KiB
TypeScript
57 lines
1.4 KiB
TypeScript
import prisma from "@/lib/prisma";
|
|
|
|
async function main() {
|
|
console.log("Sedang memulai proses seeding...");
|
|
|
|
const modelData = [
|
|
{
|
|
modelName: "Model XGBoost (Baseline)",
|
|
description:
|
|
"Model awal menggunakan parameter default XGBoost (learning_rate=0.3, max_depth=6) pada dataset yang tidak seimbang.",
|
|
accuracy: 0.71,
|
|
macroF1: 0.62,
|
|
f1Negative: 0.67,
|
|
f1Neutral: 0.34,
|
|
isActive: false,
|
|
},
|
|
{
|
|
modelName: "Model XGBoost (Tuned)",
|
|
description:
|
|
"Model dengan optimasi Hyperparameter menggunakan Grid Search untuk mencari kombinasi learning_rate dan max_depth terbaik.",
|
|
accuracy: 0.73,
|
|
macroF1: 0.66,
|
|
f1Negative: 0.69,
|
|
f1Neutral: 0.42,
|
|
isActive: false,
|
|
},
|
|
{
|
|
modelName: "Model XGBoost (Optimized)",
|
|
description:
|
|
"Model final menggunakan teknik SMOTE untuk menyeimbangkan kelas, seleksi fitur Chi-Square, dan optimasi Grid Search.",
|
|
accuracy: 0.72,
|
|
macroF1: 0.66,
|
|
f1Negative: 0.68,
|
|
f1Neutral: 0.43,
|
|
isActive: true,
|
|
},
|
|
];
|
|
|
|
for (const data of modelData) {
|
|
const model = await prisma.model.create({
|
|
data: data,
|
|
});
|
|
console.log(`Berhasil membuat model: ${model.modelName}`);
|
|
}
|
|
|
|
console.log("Proses seeding selesai.");
|
|
}
|
|
|
|
main()
|
|
.catch((e) => {
|
|
console.error(e);
|
|
process.exit(1);
|
|
})
|
|
.finally(async () => {
|
|
await prisma.$disconnect();
|
|
});
|