index.html

<!DOCTYPE html>
<html lang="en">
  <head>
    <meta charset="UTF-8" />
    <meta name="viewport" content="width=device-width, initial-scale=1.0" />
    <title>Sam Urmian</title>
    <meta
      name="description"
      content="Sam Urmian — researcher in algorithm design, automated theorem proving, and learning theory. Open-source libraries for synthetic data, adaptive learning, and graph algorithms."
    />
    <link rel="stylesheet" href="styles.css" />
  </head>
  <body>
    <main id="top">
      <header class="intro" id="about">
        <img
          class="avatar"
          src="assets/sam-urmian.jpg"
          alt="Portrait of Sam Urmian"
          width="72"
          height="72"
        />
        <div>
          <h1>Sam Urmian</h1>
          <p class="muted">Researcher at SLATE, University of Bergen</p>
        </div>
      </header>

      <nav class="site-nav" aria-label="Primary">
        <a href="#about">about</a>
        <a href="#software">software</a>
        <a href="#projects">projects</a>
        <a href="#book">book</a>
        <a href="#noki">noki</a>
      </nav>

      <section>
        <p>
          I work on theoretical computer science — learning algorithms and
          automated theorem proving — and more recently on synthetic data,
          privacy, recommender systems, and adaptive learning.
        </p>
        <p>
          PhD in Informatics, University of Bergen (advisor: Mateus de
          Oliveira Oliveira). Thesis:
          <a href="assets/sam-urmian-thesis.pdf" target="_blank" rel="noreferrer"
            ><em
              >Width-Based Dynamic Programming for Automated Theorem Proving
              in Graph Theory</em
            ></a
          >
          (PDF). MSc, Sharif University of Technology.
        </p>
        <ul class="inline-list">
          <li>
            <a
              id="email-link"
              href="#top"
              aria-label="Reveal email address"
              rel="noopener nofollow"
              ><span class="email-decoy" aria-hidden="true"
                >no-reply@example.invalid</span
              ><span class="email-prompt">email</span></a
            >
          </li>
          <li>
            <a
              id="gmail-link"
              href="#top"
              aria-label="Reveal Gmail address"
              rel="noopener nofollow"
              ><span class="email-decoy" aria-hidden="true"
                >no-reply@example.invalid</span
              ><span class="email-prompt">gmail</span></a
            >
          </li>
          <li>
            <a href="https://github.com/mlgorithm" target="_blank" rel="noreferrer"
              >github</a
            >
          </li>
          <li>
            <a
              href="https://scholar.google.com/citations?user=RktHIgcAAAAJ&hl=en"
              target="_blank"
              rel="noreferrer"
              >scholar</a
            >
          </li>
          <li>
            <a
              href="https://dblp.org/search/author?q=Farhad+Vadiee"
              target="_blank"
              rel="noreferrer"
              >dblp</a
            >
          </li>
        </ul>
      </section>

      <section id="software">
        <h2>Software</h2>
        <dl class="list">
          <dt>
            <a href="https://github.com/mlgorithm/tabdm" target="_blank" rel="noreferrer"
              >tabdm</a
            >
            <span class="tag">python</span>
          </dt>
          <dd>
            Synthetic mixed-type tabular data via diffusion in a transformed
            feature space.
          </dd>

          <dt>
            <a
              href="https://github.com/mlgorithm/orchid-ranker"
              target="_blank"
              rel="noreferrer"
              >orchid-ranker</a
            >
          </dt>
          <dd>
            Adaptive-learning recommender that picks what a learner should work
            on next.
          </dd>

          <dt>
            <a
              href="https://github.com/AutoProving/TreeWidzard-Engine"
              target="_blank"
              rel="noreferrer"
              >TreeWidzard</a
            >
          </dt>
          <dd>Engine for tree-decomposition-based algorithms.</dd>

          <dt>
            <a
              href="https://bitbucket.org/Petra2334/pace2023-code/src/master/Zygosity_Heuristic/"
              target="_blank"
              rel="noreferrer"
              >Zygosity</a
            >
            <span class="tag">pace 2023</span>
          </dt>
          <dd>Heuristic solver for computing twin-width.</dd>

          <dt>
            <a href="https://github.com/AutoProving/LUBP" target="_blank" rel="noreferrer"
              >LUBP</a
            >
          </dt>
          <dd>Learning unitary branching programs.</dd>

          <dt>
            <a
              href="https://apps.apple.com/us/search?term=Sorna%20Research"
              target="_blank"
              rel="noreferrer"
              >Sorna Research</a
            >
            <span class="tag">ios</span>
          </dt>
          <dd>Discover, follow, and save papers from public research sources.</dd>
        </dl>
      </section>

      <section id="projects">
        <h2>Projects</h2>
        <dl class="list">
          <dt>
            <a href="https://slate.uib.no/projects/aspire" target="_blank" rel="noreferrer"
              >ASPIRE</a
            >
          </dt>
          <dd>Synthetic data generation, differential privacy, federated learning.</dd>

          <dt>
            <a
              href="https://slate.uib.no/projects/artificial-intelligence-in-education-layers-of-trust-edutrust-ai"
              target="_blank"
              rel="noreferrer"
              >EduTrust AI</a
            >
          </dt>
          <dd>Trust, responsibility, and privacy in AI for education.</dd>

          <dt>
            <a
              href="https://slate.uib.no/news/slate-receives-200-million-kroner-for-research"
              target="_blank"
              rel="noreferrer"
              >AI LEARN</a
            >
          </dt>
          <dd>National AI research centre on hybrid human–AI intelligence.</dd>

          <dt>
            <a
              href="https://www.uib.no/ii/136421/prestisjeprosjekter"
              target="_blank"
              rel="noreferrer"
              >AUTOPROVING</a
            >
            <span class="tag">past</span>
          </dt>
          <dd>
            Automated theorem proving from the mindset of parameterized
            complexity theory.
          </dd>
        </dl>
      </section>

      <section id="book">
        <h2>Book</h2>
        <p>
          <a
            href="https://github.com/mlgorithm/ml-by-design"
            target="_blank"
            rel="noreferrer"
            ><strong>Machine Learning by Design</strong></a
          >
          — a draft book introducing ML from a concept-first,
          problem-solving perspective. Open on GitHub and updated as chapters
          land.
        </p>
      </section>

      <section id="noki">
        <h2>NOKI</h2>
        <p>
          I lead the Norwegian Olympiad in Artificial Intelligence, which
          trains Norway’s team for the International Olympiad in AI (IOAI).
          <a href="https://ioai.no/" target="_blank" rel="noreferrer">ioai.no</a>
        </p>
      </section>

      <footer class="muted">
        <p>Plain HTML, no trackers. <a href="#top">↑ top</a></p>
      </footer>
    </main>
    <script src="site.js"></script>
  </body>
</html>