Skip to content

Doing a Thesis in Applied Machine Learning

This repository contains guidance for students doing a bachelor's, master's, or doctoral thesis in applied machine learning supervised by Alex Jung. It applies regardless of where you are enrolled. The guide currently covers students at Aalto University (Finland) and IMC Krems University of Applied Sciences (Austria); everything that depends on your university or degree level is collected in Thesis Levels and University-Specific Information, so the rest of the guide reads the same for everyone.

Table of Contents

  1. What to Expect from Your Supervisor
  2. What Is Expected from You
  3. Thesis Levels: Bachelor's, Master's, Doctoral
  4. Getting Started
  5. Typical Timeline
  6. Recommended Development Environment
  7. Responsible Use of AI
  8. Practical Workflow
  9. Thesis Manuscript Preparation
  10. Typesetting Mathematical Texts
  11. Iterative Writing Process
  12. Self-Editing Pass (Prose Linter)
  13. Final Thesis Checklist
  14. Thesis Presentation and Self-Assessment
  15. Thesis Evaluation, Decision, and Appeals
  16. University-Specific Information
  17. References
  18. Feedback and Questions

What to Expect from Your Supervisor

As your supervisor, you can expect me to:

  • Help you clearly define your ML problem.
  • Advise on suitable ML methods, tools, and resources.
  • Be available in the weekly Thesis Garage during the teaching term (dates on the Garage page), where you can present your progress and get feedback on work in progress.
  • Guide you through thesis writing and evaluation.
  • Give feedback on complete, polished drafts and on your self-assessment.
  • Offer the opportunity to discuss your self-assessment before submission.
  • Know, or find out with you, which rules of your university apply to your thesis — but the formal process (forms, deadlines, submission system) is run by your programme, not by me.

Please note: I do not review early-stage or partial drafts — for example, a single rough chapter, an unedited first attempt, or notes that have not yet been through your own self-editing pass. I give detailed feedback only once you have a full draft that you consider finished and have run through the self-editing pass below. Sending work earlier than that does not speed things up; it means feedback gets spent on issues you could have caught yourself. (For a doctoral thesis, "full draft" applies to each article and to the summary separately; see Thesis Levels.)

What Is Expected from You

As a thesis student, you are expected to:

  • Take ownership of your research and drive progress independently.
  • Use the weekly Thesis Garage to present your progress — a result, a plot, or a problem you are stuck on. Attendance is voluntary, but it is the main place you get feedback on work in progress, so plan to come regularly.
  • Come prepared with concrete questions or results.
  • Speak up early — tell me as soon as you are stuck or falling behind schedule.
  • Use high-quality scientific references (peer-reviewed journals, reputable conferences, established textbooks).
  • Follow the writing and typesetting conventions described in this guide.
  • Know your own university's thesis regulations (registration, proposal approval, deadlines, formatting template, submission) and keep track of them yourself; see University-Specific Information.

Thesis Levels: Bachelor's, Master's, Doctoral

The writing and methodology conventions in this guide are the same at every level. What changes is the scope of the problem, the depth of the analysis, and the amount of originality expected. The table gives the typical profile; your programme's regulations define the formal requirements (credits, length, deadlines), and those always take precedence.

Bachelor's thesis (BSc) Master's thesis (MSc) Doctoral thesis (PhD)
Goal Show that you can apply established ML methods correctly to a well-defined problem and report the results in a scientific format Show that you can independently formulate an ML problem, choose and justify methods, and evaluate them critically Make an original, peer-reviewed contribution to the state of the art
Typical problem A given dataset and a clearly stated task; compare a few standard models against a sensible baseline A problem you shape yourself (often with an industry partner); design, implement, and evaluate a method, including diagnosis of failure modes A research programme spanning several related problems, each addressed in a publication
Literature review Short: the handful of papers and the textbook chapters that define your methods Systematic: position your work relative to the most relevant prior work and identify the gap you address Comprehensive and ongoing; each paper has its own related-work section, the summary ties them together
Typical duration The nominal duration set by your study programme, usually alongside courses The nominal duration set by your study programme (typically about six months full-time) 3–5 years
Where this guide applies All of it; the Self-Editing Pass matters just as much for a short text All of it All of it, applied to each article and to the summary; add the --profile paper mode of the linter suite for the articles

Bachelor's students: keep the problem small enough that you can finish the full loop — data, model, evaluation, write-up — with time left for revision. A clean, well-reported comparison of standard methods is a good bachelor's thesis; an ambitious method that is half-evaluated is not.

Master's students: the distinguishing feature is critical evaluation. Beyond reporting a test score, diagnose why a method works or fails (sensitivity analysis, error analysis, comparison to a simple baseline) and relate the findings back to your research questions.

Doctoral students: a doctoral thesis is normally article-based (a set of peer-reviewed publications plus a summary) or, less commonly, a monograph. Each paper is a self-contained piece of work with its own problem formulation, related work, and evaluation; the manuscript conventions below apply to each of them. The summary (also called the compilation part or "introductory chapter") is where this guide's advice on structure and self-contained section openers matters most, because it has to make the papers read as one argument. Doctoral supervision is offered at Aalto University; see Aalto University below for the formal requirements of the Doctoral Programme in Science.

Getting Started

To start your thesis:

  1. Formulate your ML problem clearly by identifying data points, their features, and labels (watch this video).
  2. Formulate your research questions so that each is clear, focused on a single problem, and specific about the metrics involved (guidance). Prefer "to what extent does X improve Y on metric Z" over yes/no forms like "can X beat Y", and do not bundle two questions into one. A bachelor's thesis typically has one or two research questions, a master's thesis two or three.
  3. Choose suitable ML models that you are comfortable implementing (e.g., linear regression, random forests, or neural networks).
  4. Identify data sources and evaluation criteria (e.g., test accuracy, computational efficiency).
  5. Check what your programme requires before you start — most universities require a registered topic and an approved proposal before the work formally counts. See University-Specific Information.

Detailed guidance is available in Chapter 2 of the textbook Machine Learning: The Basics (available through most university libraries) and in these lecture videos.

Typical Timeline

A thesis moves through the same phases at every level; the phases are shorter and overlap more in a bachelor's thesis, and repeat once per article in a doctoral thesis.

Phase Activities
Problem Definition Identify research question, data sources, and evaluation criteria; get the topic and proposal approved by your programme where required
Literature Review Survey related work; identify gaps your thesis addresses
Data Collection & Preprocessing Gather, clean, and explore your dataset
Modeling Implement and train ML models; run baseline experiments
Evaluation & Diagnosis Benchmarks, sensitivity analysis, error analysis
Writing Draft chapters iteratively; incorporate feedback
Self-Assessment & Presentation Write your self-assessment (see Thesis Presentation and Self-Assessment); prepare and deliver the thesis presentation or defence

Plan with the nominal duration your study programme sets for the thesis (for a master's thesis typically about six months full-time; for a doctoral thesis 3–5 years; see Thesis Levels). Your programme's deadlines, not these estimates, are binding.

Visual Studio Code is a good choice for thesis work — it supports LaTeX (via the LaTeX Workshop extension), Python, and Jupyter notebooks in one place.

The Claude Code extension for VS Code integrates an AI assistant directly into your editor. You can use it to:

  • Explain or refactor Python/LaTeX code in context
  • Get feedback on a selected paragraph or equation
  • Generate boilerplate (e.g., plotting code, pseudocode skeletons)
  • Ask questions about your codebase without leaving the editor

Note: You are accountable for everything in your thesis; an LLM cannot be. Verify every AI-generated output before you rely on it, and check that your university permits the use before you start (see Responsible Use of AI).

Responsible Use of AI

Follow your university's official policy; the rules differ between universities, and the rules for students can be stricter than those for staff (they are at IMC Krems). The links are collected under University-Specific Information. The principles below hold everywhere.

Accountability

You are accountable for the entire content of your thesis — every claim, result, and conclusion. An LLM cannot be: it cannot be examined, cannot answer for a mistake, and cannot be listed as an author. "The AI produced it" excuses nothing. Whatever tools you use, you must be able to defend every statement in your thesis as if you had written it unaided.

Disclosure

  • Disclose when and how you used AI tools in a dedicated statement; keep the Methods section for research methods. If your university's policy additionally asks for in-text marking or a note in the methods section, add that as well.
  • Record the tool, version, and settings; online services change frequently, so exact reproduction is rarely possible.
  • Include the statement even if you used no AI tools, and say so plainly — with no statement, a reader cannot tell "no AI used" from "AI used but not disclosed".
  • If your university prescribes a specific form or wording for the disclosure, use that; the dedicated statement is the minimum.

Data Protection and Intellectual Property

  • Do not upload personal, confidential, or unpublished material to public AI services — this may breach GDPR. Use local or GDPR-compliant tools for sensitive data.
  • AI output may reproduce others' work without attribution. Verify every citation independently.

Integrity-Preserving Uses

AI tools can support your work when used for:

  • Proofreading language and grammar
  • Brainstorming research directions or experiment designs
  • Explaining concepts or summarising background (always verify)
  • Generating boilerplate or plotting templates (always review and test)
  • Surfacing counterarguments to test your reasoning

Universities treat undisclosed or improper AI use as a breach of good scientific practice, with the same consequences as plagiarism — up to the thesis being declared invalid or the degree being revoked. When in doubt, ask before you use a tool, not after.

Practical Workflow

A thesis in machine learning typically involves:

  • Data Collection and Preprocessing using, e.g., pandas.
  • Model Training and Validation using, e.g., scikit-learn.
  • Model Diagnosis using numerical experiments (benchmarks, sensitivity analysis) and, when appropriate, mathematical analysis (generalisation bounds, error analysis, comparison to Bayes risk).

For academic sources, use:

  • Your university library's discovery service (access to paywalled journals and e-books; see University-Specific Information for the link)
  • IEEE Xplore: https://ieeexplore.ieee.org
  • ACM Digital Library: https://dl.acm.org
  • Scopus: https://www.scopus.com
  • Web of Science: https://www.webofscience.com
  • dblp (computer science bibliography, for finding the published version of a paper): https://dblp.org

To assess the quality of journals and conferences, consult a publication ranking such as the Finnish JUFO classification (https://jfp.csc.fi/jufoportal) or the CORE conference ranking (https://portal.core.edu.au/conf-ranks/).

If you are uncertain about a reference's quality, ask me.

Thesis Manuscript Preparation

When preparing your thesis, ensure:

  • Template: Use the official thesis template of your university and programme where one exists (see University-Specific Information); the conventions below work within any template.
  • Terminology: Use terms defined in the Aalto Dictionary of ML. You are encouraged to reuse its LaTeX source (e.g., TikZ figures).
  • Problem formulation: State clearly what the data points are and how their features and labels are defined.
  • Loss functions: Explicitly state the loss function used for training and, separately, for validation or testing.
  • Numerical results: Present and discuss results thoroughly to answer your research questions.
  • Baselines: Use appropriate baselines or benchmarks (e.g., Kaggle competitions).
  • Structure: Begin each chapter and section with an introductory paragraph explaining its content and its connection to the rest of the thesis. Where a chapter or section has subsections, the introduction should tell the reader what each subsection covers and how they fit together — stating a theme without mapping it onto the subsections leaves the reader without a map.
  • Equations: Reference all numbered equations using \eqref{}. Only number equations that are referenced in the text; leave unreferenced equations unnumbered.
  • Algorithms: Present new methods as pseudocode (see examples).
  • Figures: Ensure all figures are clear, labelled, and have informative captions (caption guidelines: Rule 4 of the PLOS "Ten Simple Rules for Better Figures"). Set the figure size to its final print width so that tick, axis, and legend text renders at no less than roughly 80% of the body-text size, and check that all plots remain readable in a grayscale printout.
  • References: Format according to IEEE guidelines unless your programme prescribes another style (IMC Krems programmes often prescribe their own; check your programme's formatting guide). Whatever the style, apply it consistently and cite the published, peer-reviewed version of a paper rather than its arXiv preprint whenever one exists (check dblp or the publisher's site).

For creating effective figures, see Edward Tufte's The Visual Display of Quantitative Information.

Typesetting Mathematical Texts

Display vs Inline Math

  • Use inline math ($...$) for short expressions within a sentence: The loss is defined as $L(\theta)$.
  • Use display math (\[ ... \] or the equation environment) for standalone equations that are central or referenced.

Punctuation with Displayed Equations

Punctuate displayed math as part of the surrounding sentence:

The empirical risk is defined as
\[
L(\theta) = \frac{1}{n} \sum_{i=1}^n \ell(f(x_i;\theta), y_i).
\]

Iterative Writing Process

  • While experimenting, keep a separate working notes file with the methods you try, the hyperparameters used, and the results — without worrying about prose, formatting, or citations. Treat it as a lab notebook.
  • Start serious writing with the literature review or methodology chapters. These settle earliest and rarely need rewriting when your results change.
  • Hold off on the results and discussion chapters until your experiments have settled — these chapters often change a lot, so writing them up too early wastes effort.
  • Write the abstract last. It is the shortest section but depends on everything else being settled.
  • Get feedback as you go through the weekly Thesis Garage. The garage — not written draft review — is where you get feedback on work in progress. Present your progress regularly: a result, a plot, a problem you are stuck on, or a short section you want reactions to. Bring specific questions. This is the fastest way to correct course early, and it is available every week of the teaching term; a finished written draft is not a prerequisite for it.
  • Also draw on feedback from peers and, where appropriate, LLM-based tools for quick checks between garage sessions.
  • Keep in mind that detailed written feedback from me comes only on a complete, polished draft you consider finished (see What to Expect from Your Supervisor). Use the weekly garage sessions for everything before that stage.
  • Expect and budget for multiple revision rounds before submission. Some programmes (for example the master's programmes at IMC Krems) formally allow a rejected thesis to be sent back for revision only a limited number of times, so the revision rounds should happen before the official submission, not after.

Self-Editing Pass (Prose Linter)

Once a draft is near-final, read it through once looking only for the recurring prose defects below. Each is cheap to fix yourself and distracting for a reader — or your supervisor — to flag repeatedly. Fixing them before review means feedback can focus on the substance of your work rather than on mechanics.

  • Excessive forward referencing — "as we will see in Chapter 5" scattered throughout. Present material where the reader needs it; a few pointers are fine, but a habit of them signals disorganised structure. The linter suite flags forward-looking cue phrases (prose_lint.py), floats referenced long before they appear (crossref_forward_lint.py), and concepts used before they are defined (forward_ref_lint.py / forward_ref_lint_llm.py) — run these as a first pass.
  • Undefined or re-defined acronyms — expand each acronym at first use, then use it consistently; do not re-expand it later.
  • Inconsistent terminology — pick one term per concept (following the Aalto Dictionary of ML) and do not silently switch between synonyms (e.g., "data point" vs "sample" vs "instance").
  • Vague quantifiers — "significantly", "very", "a lot" without a number. Give the figure or drop the word.
  • Uncited claims — every empirical or historical assertion that is not your own result needs a citation.
  • Dangling references — "this shows", "it follows" where "this" or "it" has no clear antecedent.
  • Unmotivated sections — a chapter or section that opens without a sentence stating what it covers and why it belongs here.
  • Openers that lean on the previous section — readers (and examiners) enter your thesis at headings, so the first sentence after a heading must stand alone. "A difference of almost an order of magnitude illustrates diverging views" — a difference in what? Name the antecedent ("Published estimates of tracking's revenue benefit differ by almost an order of magnitude ...") or the opener is opaque to anyone who did not just read the preceding section. Named references ("the harms reviewed in Section 2.2.4") are fine; bare pronouns and unexplained noun phrases pointing across the heading are not. The same applies within a section: consecutive paragraphs must not jump topic without a bridging sentence.
  • Jargon and undefined evaluative claims — "smoothest convergence", "performs well", "training is stable": define and measure the property, or drop the claim. A nearby number (a hyperparameter value, a sample size) does not quantify the claim itself. Beware evaluative use of words with a precise technical meaning ("smooth" means differentiable).
  • "Statistically significant" — use only with an actual statistical test. Without one the phrase claims evidence you do not have; write "reduces sampling variability" or report the spread instead.
  • Type, range, and dimension mismatches in formal claims — a relation must relate objects of the same kind. An active-constraint set cannot "approximate" a strategy (a bound-assignment vector) unless you first state the map between them; a parameter you defined as a probability cannot be set to 10; a dimensionless speedup is not measured in seconds. State the correspondence, keep every value inside its defined range, and drop stray units. The type_consistency_lint_llm.py linter flags these as TYPE-MISMATCH, RANGE, and DIMENSION (and BRIDGE-LOOSE when a defined map licenses the shorthand — restate the claim over the mapped object).
  • Tense drift — keep the tense consistent within a passage when describing methods and results.

Reading aloud, or reading one defect type at a time from start to finish, catches far more than a single general pass.

A linter suite covering nearly every defect on this list (all but the statistical-significance check) — and most of the manuscript checklist above — is available in assets/linters/: run python3 run_all_linters.py thesis.pdf on your compiled PDF (or on your LaTeX sources) before every revision round. For the fast pass the scripts need only Python 3 plus pdftotext or PyMuPDF; see the linter README. The --llm checks default to the Aalto AI API (Aalto network or VPN), but can also run against a local on-device model (e.g. mlx_lm.server on Apple Silicon, or Ollama) or any OpenAI-compatible endpoint your own university provides, so your unpublished draft never leaves your machine or your institution — see Data handling in the README. Students outside Aalto should use one of those alternatives.

Final Thesis Checklist

Before submitting, verify each item below. Links point to the relevant section of this guide.

  • ML problem is precisely formulated: data points, features, and labels are clearly defined (see Getting Started)
  • Loss functions for training and evaluation are explicitly stated (see Manuscript Preparation)
  • Methods are clearly described, including pseudocode for new algorithms
  • Baselines or benchmarks are included and discussed
  • All figures have labelled axes and informative captions
  • All numbered equations, tables, and figures are referenced in the text
  • Citations are formatted consistently according to IEEE guidelines or the style your programme prescribes
  • Acronyms are expanded at first use and terminology is consistent (see Self-Editing Pass)
  • A prose self-editing pass has been done for forward referencing, vague quantifiers, and dangling references (see Self-Editing Pass)
  • Research questions are focused, specific, and not answerable with a bare yes/no (see Getting Started)
  • Solver/library versions, key tolerance values, and hardware are reported for all experiments
  • Figure text is legible at print size and in grayscale
  • The AI-use statement is included, naming tools and versions (also if no AI tools were used)
  • The thesis uses your university's official template and meets its formal requirements (see University-Specific Information)
  • Your self-assessment is written (Aalto master's students: fill in the School of Science assessment statement template; see below)

Thesis Presentation and Self-Assessment

After completing the thesis manuscript:

  • Write a self-assessment with explicit references to sections of your thesis. Aalto master's students fill in the School of Science assessment statement template — the form examiners use — as their self-assessment; this is my practice, not a programme requirement. Students elsewhere write a one-page self-assessment along the same lines. It is the basis for our discussion in any case.
  • Read the grading criteria of your programme before submission so you know what a high-quality thesis looks like in your examiners' eyes. For Aalto master's theses this is the School of Science grade characterisation PDF; for other programmes see University-Specific Information.
  • Optionally, request a meeting to discuss your self-assessment before submission.
  • Prepare your thesis presentation — either live during a garage session or as a recorded video (see examples). Where your programme requires a formal presentation or oral defence (for example the kommissionelle Gesamtprüfung at IMC Krems or a doctoral defence), the garage is a good place to rehearse it.

Thesis Evaluation, Decision, and Appeals

YouTube overview (recorded for Aalto master's students; the principles are the same elsewhere)

How the grade is decided

  • Your thesis is evaluated against the official criteria of your programme and university, not against this guide. This guide tells you how to meet those criteria.
  • I prepare a written evaluation and, where the procedure provides for one, a grade proposal based on these criteria.
  • The final decision is made by your programme or school following its formal procedure (a degree committee, a programme director, an examination board).
  • Study your programme's grading criteria carefully before submission (links under University-Specific Information).

Transparency and feedback

  • You have the right to see the evaluation criteria applied to your thesis and to receive the written evaluation.
  • You may request clarification on how your thesis was assessed and how the grade was formed.
  • Your self-assessment is an important part of this process.

Appealing a grading decision

  • If you believe an error occurred, you can request a review of the decision. The exact instrument differs by country: at Aalto it is a request for rectification of the grade, at IMC Krems a complaint about a defect in how a negatively assessed examination was conducted. The deadlines are short — two weeks at both universities — so act promptly.
  • The formal procedures are linked under University-Specific Information.

Practical advice: If unsure whether an appeal is appropriate, discuss the evaluation with your supervisor first. Many issues can be resolved without a formal appeal.

University-Specific Information

Everything in this guide that depends on where you are enrolled is collected here. Your university's regulations always take precedence over this guide. If your university is not listed, send me the link to its thesis regulations and I will add a section.

Aalto University

Theses are supervised in the Department of Computer Science (School of Science, SCI).

Where to look
Bachelor's thesis Instructions and the evaluation rubric on the programme's thesis page; templates and the seminar schedule on the JOIN.bsc course page in MyCourses; past theses in Aaltodoc
Master's thesis SCI Master's Thesis Guide on MyCourses — templates, forms, evaluation rules; start with the "Get oriented" section
Doctoral thesis Doctoral thesis in the Aalto Doctoral Programme in Science — article-based vs monograph, pre-examination, defence; general information at Aalto Doctoral Education
Grading criteria (MSc) The School of Science (SCI) grade characterisation, which assumes a six-month target time for the thesis process, and the assessment statement template examiners use, which I ask you to fill in as your self-assessment
Peer-review forms Review forms from my Aalto courses, useful as a model for reviewing your own draft: Machine Learning project, Federated Learning project
Responsible AI use Tips for using AI for students; for doctoral students also Responsible use of generative AI in the research process, which is written for researchers
Library Aalto University Library (Primo); the Finnish JUFO ranking for publication quality
Appeals Academic appeals at Aalto University: a request for rectification must be submitted within 14 days of being notified of the grade
LLM access for the linters The linter suite defaults to the Aalto AI API / LLM Gateway (Aalto network or VPN)

IMC Krems University of Applied Sciences

IMC Krems (IMC Fachhochschule Krems, Austria) offers bachelor's and master's programmes; as a university of applied sciences it does not award doctorates. The binding rules are in the university's Studien- und Prüfungsordnung (study and examination regulations); the detailed thesis guides it refers to are internal documents available to enrolled students on the eDesktop intranet.

Where to look
Regulations Satzungsteil Studien- und Prüfungsordnung (German; FHR-5-0035). Bachelor's theses: section 3.8; master's theses: section 3.9
Thesis guides Leitfaden für Bachelorarbeiten und Bachelorprüfungen (FHM-5-0008), Leitfaden für Masterarbeiten und Masterprüfungen (FHR-5-0009), and the formatting guide Leitfaden für die formale Gestaltung schriftlicher/wissenschaftlicher Arbeiten (FHM-5-0003) — all on the eDesktop; ask your programme's study services if you cannot find them
Process The topic and the written proposal (Exposé) must be approved by the programme director; approval is a prerequisite for submitting the thesis, and surveys or data collection may only begin after it (following the IMC IT policy and data-protection rules). A master's thesis can be submitted for approval at the earliest three months after the proposal was approved; topics involving ethics-relevant data additionally need ethics-committee approval, which the student applies for
Assessment The supervisors write a moderated assessment report (Gutachten), which is handed to you together with the grade. A master's thesis that is not approved can be returned for revision at most twice. The thesis is followed by a presentation and oral examination (kommissionelle Gesamtprüfung)
Good scientific practice and AI Section 3.7 of the regulations: plagiarism, ghostwriting, data falsification, and improper use of AI invalidate the thesis. Students are bound by the student AI guideline Richtlinie zur Verwendung von KI durch Studierende (FHR-5-004, on the eDesktop), which is in part stricter than the general staff guideline
Library The Campus Library, run by the University for Continuing Education Krems and open to the public, serves IMC students; the databases IMC licenses are reached through the eDesktop
Appeals The assessment itself cannot be appealed, but a complaint about a defect in how a negatively assessed examination was conducted can be filed with the programme director within two weeks; you may inspect your assessment documents within six months of the grade being announced (regulations, sections 3.1 and 3.2; legal basis: Fachhochschulgesetz)
LLM access for the linters Run the --llm linters against a local model or an endpoint your university provides; the Aalto endpoints require an Aalto account (see Data handling)

Other universities

Theses supervised earlier at TU Wien are listed in theses.md. If you are enrolled elsewhere and would like me to co-supervise, the general parts of this guide apply unchanged; send me your university's thesis regulations with your first email.

References

ML Fundamentals

  • A. Jung, Machine Learning: The Basics. Singapore: Springer, 2022. [DOI]
  • A. Jung et al., "The Aalto Dictionary of Machine Learning," Aalto University. [GitHub]
  • S. Shalev-Shwartz and S. Ben-David, Understanding Machine Learning: From Theory to Algorithms. Cambridge University Press, 2014.

Writing and Typesetting

Feedback and Questions

Reach out via: