EARLY REGISTRATION

৳10,000 OFF Regular fee ৳50,000 · Early registration ৳40,000

STARTING JANUARY 2027

UPCOMING COURSE · 2027

AI & Machine Learning Research Bootcamp

A research-first, hands-on program designed to take learners from research foundations and literature review to dataset development, AI/ML experimentation, proposed model development, rigorous evaluation, scientific writing, and publication-ready research.

Research Foundation AI & Machine Learning Hands-on Research Proposed Models Foundation Models Scientific Writing Publication

Months

Class Per Week

Research Project

01+

Publication Ready Target

Weekly class timing is planned around student availability. Sessions are generally 1 to 3 hours and may continue longer when research implementation, experiments, discussion, or hands-on work requires additional time.

PROGRAM SNAPSHOT

Learn AI/ML through the research lifecycle

The program focuses on how research is actually conducted, from identifying a problem and reviewing literature to building, evaluating, writing, and preparing research for publication.

What students will learn

Research methodology and scientific thinking
Literature search, synthesis and research gap analysis
Research and AI ethics
Dataset collection, processing and annotation
Python and reproducible AI/ML workflows
Machine learning and deep learning experimentation
Transformers, Hugging Face and foundation models
Proposed model design and implementation
Rigorous model evaluation and statistical analysis
Scientific writing and AI-assisted research workflows
Publication strategy and peer review

Hands-on research experience

Research problem selection
Literature matrix development
Dataset audit and annotation
Baseline model implementation
Proposed model implementation
Experiment tracking
Ablation and comparative experiments
Error and failure analysis
Research manuscript development
Publication preparation

WHY AMIR ACADEMY

Built around research, not just lectures

Every major stage combines learning, practical work, feedback, assignments, and a concrete research deliverable.

Research First

Learn how to turn a research problem into a structured and testable AI/ML study.

Hands-on Work

Build datasets, implement models, run experiments and analyze real results.

Research Deliverables

Each module produces an assignment, artifact, report, experiment or research document.

Proposed Models

Learn how to move from a research gap to a proposed architecture and implementation.

Rigorous Evaluation

Understand which metrics and validation strategies matter for different research problems.

Publication Focus

Work toward a publication-ready manuscript through structured research and writing.

RESEARCH METHODOLOGY

From research question to publication

The hands-on methodology follows a complete research workflow. Students repeatedly apply the concepts through assignments, implementation and research deliverables.

01 Problem

›

02 Literature

›

03 Research Gap

›

04 Dataset

›

05 Baseline

›

06 Proposed Model

›

07 Experiments

›

08 Evaluation

›

09 Writing

›

10 Publication

COMPLETE CURRICULUM

14 Research-Focused Modules

Click any module to explore the topics, assignment and expected deliverables.

01 Research Foundations Build the mindset and methodology required for research. ＋

Research methodology

Scientific thinking

Research problems and questions

Research objectives

Research hypotheses

Experimental design

Research planning

Selecting AI/ML research directions

Assignment · Research Problem Brief

Select a potential AI/ML research area.
Define the problem and research questions.
Draft objectives and preliminary hypotheses.

Deliverable · Research Foundation Document

Problem statement
Research questions
Objectives and initial hypothesis

02 Literature Review & Research Gap Learn to find, read, compare and synthesize academic literature. ＋

Academic literature search

Google Scholar

Research databases

Critical paper reading

Literature matrix

Literature synthesis

Systematic literature review fundamentals

Research gap identification

Novelty analysis

Research hypotheses

Assignment · Literature Investigation

Find relevant peer-reviewed research papers.
Compare methods, datasets and results.
Identify limitations and unexplored areas.

Deliverable · Literature Matrix

Structured literature matrix
Research gap statement
Preliminary novelty direction

03 Research Ethics, AI Ethics & Academic Integrity Conduct research responsibly and maintain academic integrity. ＋

Research ethics

AI ethics

Fairness and bias

Privacy and responsible data use

Explainability

Research integrity

Plagiarism

Authorship and contributorship

Author order

Conflict of interest

Publication ethics

Responsible Generative AI use

Assignment · Ethics Review

Analyze an AI research scenario.
Identify ethical and integrity risks.
Propose responsible research practices.

Deliverable · Research Ethics Checklist

Ethical risk assessment
Authorship plan
Responsible AI checklist

04 Data Collection, Processing & Annotation Build reliable datasets for AI/ML research. ＋

Research data sources

Data collection

Dataset design

Data cleaning

Preprocessing

Data annotation

Annotation guidelines

Quality control

Class imbalance

Data leakage

Train/validation/test design

Dataset documentation

Assignment · Dataset Audit

Inspect a research dataset.
Identify missing values, imbalance and leakage risks.
Design an annotation or preprocessing workflow.

Deliverable · Data Audit Report

Dataset profile
Data quality findings
Preprocessing and annotation plan

05 Python, Anaconda & AI-Assisted Coding Build reproducible coding workflows for AI/ML research. ＋

Python for research

Anaconda

Environment management

Jupyter Notebook

NumPy

Pandas

Matplotlib

Scikit-learn

TensorFlow

PyTorch

AI-assisted coding

AI-assisted debugging

Assignment · Reproducible Research Notebook

Create a clean research environment.
Load and inspect a dataset.
Perform preprocessing and visualization.

Deliverable · Research Codebase

Structured notebook
Environment specification
Reproducible preprocessing workflow

06 Kaggle, Colab & Research Computing Learn to use computational resources efficiently. ＋

Kaggle datasets

Kaggle notebooks

Google Colab

CPU fundamentals

GPU fundamentals

TPU fundamentals

GPU memory

Batch optimization

Mixed precision

Efficient data loading

Training optimization

Computational cost

Assignment · Compute Optimization

Run an experiment using available compute resources.
Compare resource and batch configurations.
Measure training efficiency.

Deliverable · Compute Experiment Report

Resource comparison
Runtime and memory observations
Optimization recommendations

07 Machine Learning & Deep Learning Establish strong modeling and experimentation foundations. ＋

Supervised learning

Unsupervised learning

Classification

Regression

Feature engineering

Baseline models

Cross-validation

Hyperparameter optimization

Ensemble learning

Neural networks

CNNs

Attention mechanisms

Assignment · Baseline Benchmark

Implement multiple baseline models.
Use appropriate validation methodology.
Compare performance and computational cost.

Deliverable · Baseline Benchmark Report

Model comparison table
Validation results
Baseline selection rationale

08 Transformers, Hugging Face & Foundation Models Explore modern pretrained and foundation model workflows. ＋

Transformer architecture

Attention

Transfer learning

Pretrained models

Hugging Face ecosystem

Model Hub

Dataset Hub

Tokenizers

Fine-tuning

Parameter-efficient fine-tuning

Large Language Models

Vision and multimodal models

Assignment · Foundation Model Experiment

Select an appropriate pretrained model.
Run inference or fine-tuning.
Analyze suitability for a research task.

Deliverable · Foundation Model Evaluation

Model selection rationale
Experiment results
Research application assessment

09 Proposed Model Development Move from research gap and novelty to a proposed methodology. ＋

From gap to methodology

Research novelty

Model architecture design

Model components

Feature representation

Feature fusion

Attention mechanisms

Custom loss functions

Architecture diagrams

Baseline vs proposed model

Proposed model implementation

Methodological contribution

Assignment · Proposed Model Design

Design a model addressing the identified research gap.
Explain each proposed component.
Define the expected methodological contribution.

Deliverable · Proposed Methodology

Architecture diagram
Methodology description
Novelty and contribution statement

10 Hands-on AI/ML Research Project Implement the research pipeline from data to proposed model. ＋

Research problem formulation

Dataset preparation

Annotation and preprocessing

Baseline implementation

Proposed model implementation

Training and optimization

Experiment tracking

Comparative experiments

Ablation studies

Error analysis

Failure analysis

Reproducibility

Assignment · Research Implementation

Implement the selected research pipeline.
Train baseline and proposed models.
Record experiments and analyze failures.

Deliverable · Research Experiment Package

Working research code
Experiment logs
Baseline and proposed model results

11 Model Evaluation & Reliability Learn which evaluation method matters for which research problem. ＋

Accuracy

Precision

Recall

F1-score

ROC-AUC

PR-AUC

Confusion matrix

MCC

Cohen's Kappa

MAE, MSE and RMSE

R²

Cross-validation

Statistical significance

Confidence intervals

Calibration

Error analysis

Assignment · Evaluation Strategy

Select metrics for a given research problem.
Evaluate a trained model using multiple measures.
Explain why each metric is relevant.

Deliverable · Evaluation Report

Metric selection rationale
Evaluation table and figures
Error and reliability analysis

12 Overfitting, Underfitting & Optimization Diagnose generalization problems and improve model reliability. ＋

Bias and variance

Overfitting

Underfitting

Learning curves

Training vs validation performance

Regularization

Dropout

Weight decay

Early stopping

Data augmentation

Hyperparameter optimization

Data leakage

Generalization

Robustness

GPU optimization

Memory efficiency

Assignment · Model Diagnosis

Analyze training and validation curves.
Identify overfitting or underfitting.
Apply appropriate optimization strategies.

Deliverable · Optimization Report

Diagnosis evidence
Optimization experiments
Before and after comparison

13 Scientific Writing & AI-Assisted Paper Writing Convert research results into a structured scientific manuscript. ＋

Research paper structure

Abstract

Introduction

Literature review

Methodology

Proposed model section

Experimental setup

Results

Discussion

Limitations

Future work

References

Citation management

AI-assisted writing

AI-assisted literature organization

Fabricated citation detection

Assignment · Manuscript Section

Draft selected sections of the research paper.
Use AI tools responsibly for editing and organization.
Verify every AI-assisted factual claim and reference.

Deliverable · Manuscript Draft

Structured manuscript sections
Research figures and tables
Verified citations and references

14 Publication & Research Career Prepare research for submission and build a stronger academic profile. ＋

Journal selection

Conference selection

Journal scope

Q1, Q2, Q3 and Q4

Predatory journals

Submission workflow

Peer review

Reviewer response

Revision strategy

Cover letters

Publication strategy

Research profile development

Assignment · Publication Strategy

Identify suitable publication venues.
Compare scope, indexing and suitability.
Prepare a submission strategy.

Deliverable · Publication Package

Target journal/conference shortlist
Submission checklist
Publication roadmap

HANDS-ON CAPSTONE

One complete AI/ML research project

Students work through a structured research project that connects the modules into one complete workflow.

01 · Research Problem

Define a focused problem, research questions, objectives, hypotheses and expected contribution.

02 · Literature

Review previous work, compare methods and identify the research gap.

03 · Dataset

Collect or select data, perform preprocessing, annotation, quality control and leakage analysis.

04 · Baselines

Establish credible baseline models and define an appropriate experimental protocol.

05 · Proposed Model

Design and implement a proposed methodology addressing the identified research gap.

06 · Evaluation

Conduct comparative experiments, ablations, statistical analysis and failure analysis.

07 · Writing

Develop the methodology, results, discussion and other sections of a scientific manuscript.

08 · Publication

Prepare the research package and identify suitable publication venues.

EVALUATION MATTERS

Learn when each metric matters

Model evaluation is not about reporting as many metrics as possible. Students learn how the research problem, class distribution, decision objective and experimental design determine the appropriate evaluation strategy.

Balanced Classification

Accuracy, Precision, Recall, F1 and confusion matrix can provide complementary evidence.

Imbalanced Classification

PR-AUC, Recall, F1, MCC and class-specific analysis may become more informative than accuracy alone.

Ranking & Discrimination

ROC-AUC and PR-AUC help assess ranking and discrimination across thresholds.

Regression

MAE, MSE, RMSE and R² provide different views of prediction error and explained variance.

Reliability

Calibration, confidence intervals and statistical analysis help assess reliability beyond a single score.

Generalization

Cross-validation, external validation, error analysis and failure analysis help assess robustness.

SELECTED OPPORTUNITIES

What may come after the program

Strong-performing students may have opportunities to continue their research journey beyond the classroom, subject to project quality, availability and AMIR Lab selection.

Research Internship Opportunity

Selected students may have an opportunity to pursue research internship work with AMIR Lab based on research performance, project quality and available positions.

APC Support Opportunity

Selected high-quality research projects may have an opportunity to receive APC support from AMIR Lab, subject to research quality, target venue, publication suitability and AMIR Lab approval.

RESEARCH TRACK RECORD

40+ 2025 Batch Alumni

A research-oriented learning environment

More than 40 alumni came through the 2025 batch. Many alumni went on to publish research papers after completing their research journey, reflecting the practical and research-focused direction of the program.

PROGRAM INVESTMENT

Regular Fee · ৳50,000

৳40,000

EARLY REGISTRATION · SAVE ৳10,000

Program starts in January 2027. Early registration price is currently ৳40,000.

৳10,000+ First Payment

03 Installments

06 Months

Program highlights

6-month research-focused program
1 student-friendly live class every week
Hands-on AI/ML research project
Assignments after every major module
Research deliverables throughout the program
Proposed model development and implementation
Publication-ready manuscript target
Potential AMIR Lab APC support opportunity
Potential AMIR Lab research internship opportunity

Research. Build. Evaluate. Publish.

Start building the research skills, technical foundation and publication workflow needed for serious AI/ML research.

UPCOMING · JANUARY 2027