HendriKarisma

← Teaching class

Machine LearningMachine Learning

MITI.302 · Ganjil 2026/2027

Introductory machine learning course at STMIK Tazkia: AI landscape, inductive learning, classical models (Mitchell), then optimisation, neural networks, deep learning, transformers and LLM.Mata kuliah pengantar machine learning di STMIK Tazkia: peta AI, pembelajaran induktif, model klasik mengikuti buku Tom Mitchell, lalu optimasi, jaringan saraf, deep learning, transformer, dan pengenalan LLM.

Materials on this page are open. Assignments and quizzes belong to the class — log in to see them.Materi di halaman ini terbuka. Tugas dan kuis milik kelas — masuk dulu untuk melihatnya.
1

Machine Learning 101: peta AI, kelas masalah, definisi belajarMachine Learning 101: peta AI, kelas masalah, definisi belajar

Opening meeting: you map the field of AI and where machine learning sits in it, distinguish problem classes (P, NP) and complexity notation, analytic vs numerical solutions, and Mitchell's definition of learning (T, P, E).Pertemuan pembuka: Anda memetakan AI dan posisi machine learning di dalamnya, membedakan kelas masalah (P, NP) dan notasi kompleksitas, solusi analitik vs numerik, serta definisi belajar menurut Mitchell (T, P, E).

MaterialsMateri
2

Concept learning, Find-S, version space, dan bias induktifConcept learning, Find-S, version space, dan bias induktif

You model learning as a search through a hypothesis space, run the Find-S and Candidate-Elimination algorithms, and understand version spaces and inductive bias.Anda memodelkan pembelajaran sebagai pencarian pada ruang hipotesis, menjalankan algoritma Find-S dan Candidate-Elimination, serta memahami version space dan bias induktif.

MaterialsMateri
3

Pohon keputusan: entropy, information gain, ID3, C4.5/J48Pohon keputusan: entropy, information gain, ID3, C4.5/J48

You build decision trees using entropy and information gain (ID3, C4.5), prune them, and explain what causes overfitting.Anda membangun pohon keputusan dengan entropy dan information gain (ID3, C4.5), memangkasnya (pruning), dan menjelaskan penyebab overfitting.

MaterialsMateri
4

Ensemble: bagging, random forest, boosting, XGBoostEnsemble: bagging, random forest, boosting, XGBoost

You learn and apply ensemble methods (bagging, random forest, boosting, XGBoost) to improve model accuracy and stability.Anda mempelajari dan menerapkan metode ensembel (bagging, random forest, boosting, XGBoost) untuk meningkatkan akurasi dan kestabilan model.

MaterialsMateri
5

Bayesian learning, Naive Bayes, GMM, dan EMBayesian learning, Naive Bayes, GMM, dan EM

You evaluate hypotheses statistically (confidence intervals, k-fold cross-validation) and apply Bayesian learning (Bayes' theorem, MAP, maximum likelihood).Anda mengevaluasi hipotesis secara statistik (selang kepercayaan, k-fold cross validation) dan menerapkan pembelajaran Bayesian (teorema Bayes, MAP, maximum likelihood).

MaterialsMateri
6

kNN dan K-MeanskNN dan K-Means

You apply naive Bayes and support vector machines (margin, kernels, the C and gamma parameters) and explain the assumptions behind each.Anda menerapkan naive Bayes dan support vector machine (margin, kernel, parameter C dan gamma) serta menjelaskan asumsi masing-masing.

MaterialsMateri
7

SVM dan Hidden Markov ModelSVM dan Hidden Markov Model

You apply instance-based learning (kNN) and linear regression, then compare them with the earlier models.Anda menerapkan pembelajaran berbasis instance (kNN) dan regresi linier, lalu membandingkannya dengan model-model sebelumnya.

MaterialsMateri
UTS

UTS - tugas besar: penerapan model klasik pada masalah nyataUTS - tugas besar: penerapan model klasik pada masalah nyata

Midterm exam as a major project: apply the material of meetings 1-7 to one real problem, from problem formulation and data preparation to choosing a classical model, evaluation and reporting.Ujian Tengah Semester berupa tugas besar: menerapkan materi pertemuan 1-7 pada satu masalah nyata, dari perumusan masalah, penyiapan data, pemilihan model klasik, sampai evaluasi dan laporan.

MaterialsMateri
9

Regresi linier, evaluasi hipotesis, dan gradient descentRegresi linier, evaluasi hipotesis, dan gradient descent

You treat learning as optimization: loss functions, gradient descent, SGD and mini-batches on linear regression, ridge/lasso regularization, and logistic regression.Anda memandang pembelajaran sebagai optimasi: fungsi loss, gradient descent, SGD dan mini-batch pada regresi linier, regularisasi ridge/lasso, hingga regresi logistik.

MaterialsMateri
10

Perceptron, multilayer perceptron, dan backpropagationPerceptron, multilayer perceptron, dan backpropagation

You study the perceptron, the multilayer perceptron, activation functions, and the backpropagation algorithm as an application of the chain rule.Anda mempelajari perceptron, multilayer perceptron, fungsi aktivasi, dan algoritma backpropagation sebagai penerapan aturan rantai.

MaterialsMateri
11

Melatih jaringan saraf: optimizer, regularisasi, diagnosisMelatih jaringan saraf: optimizer, regularisasi, diagnosis

You train neural networks stably (initialization, the Adam optimizer, batch normalization, dropout, early stopping) and diagnose training problems from loss curves.Anda melatih jaringan saraf secara stabil (inisialisasi, optimizer Adam, batch normalization, dropout, early stopping) dan mendiagnosis masalah pelatihan dari kurva loss.

MaterialsMateri
12

Deep neural network dan transfer learningDeep neural network dan transfer learning

You study deep neural network architectures, why depth helps, the vanishing gradient problem, residual connections, and transfer learning.Anda mempelajari arsitektur deep neural network, alasan kedalaman membantu, masalah vanishing gradient, residual connection, dan transfer learning.

MaterialsMateri
13

Convolutional neural networkConvolutional neural network

You study and apply convolutional neural networks to image data: convolution, pooling, classic architectures (LeNet to ResNet), and transfer learning.Anda mempelajari dan menerapkan convolutional neural network untuk data citra: konvolusi, pooling, arsitektur klasik (LeNet sampai ResNet), dan transfer learning.

MaterialsMateri
14

Attention dan arsitektur transformerAttention dan arsitektur transformer

You study the attention mechanism and the transformer architecture: self-attention, multi-head attention, positional encoding, and encoder and decoder blocks.Anda mempelajari mekanisme attention dan arsitektur transformer: self-attention, multi-head attention, positional encoding, serta blok encoder dan decoder.

MaterialsMateri
15

Large language model, generative AI, dan etikanyaLarge language model, generative AI, dan etikanya

You learn how large language models and generative AI work (pretraining, fine-tuning, prompting, RAG) and weigh their limitations and ethics; followed by group project presentations.Anda mempelajari dasar kerja large language model dan generative AI (pretraining, fine-tuning, prompting, RAG) serta keterbatasan dan etikanya; dilanjutkan presentasi proyek kelompok.

MaterialsMateri
UAS

UAS - tugas besar: penerapan jaringan saraf pada masalah nyataUAS - tugas besar: penerapan jaringan saraf pada masalah nyata

Final exam as a major project: apply the material of meetings 9-15 to one real problem, covering model design, training, evaluation, and a discussion of limitations.Ujian Akhir Semester berupa tugas besar: menerapkan materi pertemuan 9-15 pada satu masalah nyata, mencakup perancangan model, pelatihan, evaluasi, dan pembahasan keterbatasan.

MaterialsMateri