<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>News &amp; Updates |</title><link>https://naram-m.github.io/blog/</link><atom:link href="https://naram-m.github.io/blog/index.xml" rel="self" type="application/rss+xml"/><description>News &amp; Updates</description><generator>HugoBlox Kit (https://hugoblox.com)</generator><language>en-us</language><lastBuildDate>Thu, 01 Oct 2026 00:00:00 +0000</lastBuildDate><image><url>https://naram-m.github.io/media/icon.svg</url><title>News &amp; Updates</title><link>https://naram-m.github.io/blog/</link></image><item><title>Paper Accepted at NeurIPS 2026</title><link>https://naram-m.github.io/blog/neurips-2026/</link><pubDate>Thu, 01 Oct 2026 00:00:00 +0000</pubDate><guid>https://naram-m.github.io/blog/neurips-2026/</guid><description>&lt;p&gt;Our paper &lt;em&gt;&amp;ldquo;Dynamic Regret in Online Convex Optimization with Indicator Switching Costs,&amp;rdquo;&lt;/em&gt; with George Iosifidis, has been accepted at &lt;strong&gt;NeurIPS 2026&lt;/strong&gt;.&lt;/p&gt;
&lt;p&gt;The paper studies online decisions that incur a fixed cost whenever they change, capturing startup overheads such as server activation, model deployment, and cache updates. Existing guarantees for such costs only compete against a fixed benchmark. We propose a meta-learning algorithm whose dynamic regret is minimax-optimal, up to logarithmic factors, for tracking piecewise-constant comparators, without prior knowledge of how often the comparator changes.&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;&lt;strong&gt;Preprint:&lt;/strong&gt;
📄&lt;/li&gt;
&lt;/ul&gt;</description></item><item><title>Successfully Defended My Ph.D. Dissertation</title><link>https://naram-m.github.io/blog/phd-defense/</link><pubDate>Mon, 24 Nov 2025 00:00:00 +0000</pubDate><guid>https://naram-m.github.io/blog/phd-defense/</guid><description>&lt;p&gt;On &lt;strong&gt;24 November 2025&lt;/strong&gt;, I successfully defended my Ph.D. dissertation,
&lt;em&gt;&amp;ldquo;Optimistic Learning with Applications to Caching Networks,&amp;rdquo;&lt;/em&gt; at Delft University of Technology.&lt;/p&gt;
&lt;p&gt;The work was carried out in the
,
under the supervision of &lt;strong&gt;Prof. dr. K. G. Langendoen&lt;/strong&gt; (promotor) and &lt;strong&gt;Dr. G. Iosifidis&lt;/strong&gt; (copromotor).&lt;/p&gt;
&lt;style&gt;
/* Folded "thesis at a glance" panel: closed by default, opens on click. */
.thesis-glance {
margin: 2rem 0;
padding: 0.75rem 1rem;
border: 1px solid color-mix(in srgb, currentColor 15%, transparent);
border-left: 3px solid var(--color-primary-500, #3b82f6);
border-radius: 0 0.5rem 0.5rem 0;
}
.thesis-glance summary {
cursor: pointer;
font-weight: 600;
color: var(--color-primary-600, #2563eb);
}
.dark .thesis-glance summary { color: var(--color-primary-300, #93c5fd); }
.thesis-glance[open] summary { margin-bottom: 0.75rem; }
.thesis-figure { margin: 1.5rem 0 0.5rem; }
.thesis-figure img { display: block; width: 100%; max-width: 26rem; height: auto; margin: 0 auto; }
.thesis-figure figcaption { font-size: 0.85em; text-align: center; margin-top: 0.75rem; }
/* Transparent vector figure: in dark mode invert the ink and restore the hues. */
.dark .thesis-figure img { filter: invert(1) hue-rotate(180deg); }
&lt;/style&gt;
&lt;details class="thesis-glance"&gt;
&lt;summary&gt;The thesis at a glance&lt;/summary&gt;
&lt;p&gt;The thesis asks how sequential decision-making algorithms can leverage untrusted predictions while still maintaining worst-case guarantees. Its central application is caching in content delivery networks, where the proposed algorithms are validated experimentally on multiple established datasets and benchmarks. Each chapter addresses this question under a progressively richer benchmark: static regret for coded caching, α-static regret for whole-file caching, dynamic regret for tracking a moving target, and policy regret for systems with memory.&lt;/p&gt;
&lt;figure class="thesis-figure"&gt;
&lt;img src="thesis-progression.svg" alt="Technical progression of the thesis chapters: optimistic learning under static, alpha-static, dynamic, and policy regret, with the actions and comparator of each chapter." loading="lazy"&gt;
&lt;figcaption&gt;Technical progression of the chapters. Each chapter adopts a distinct regret metric suited to its actions set and choice of comparator(s).&lt;/figcaption&gt;
&lt;/figure&gt;
&lt;/details&gt;
&lt;h2 id="materials"&gt;Materials&lt;/h2&gt;
&lt;ul&gt;
&lt;li&gt;&lt;strong&gt;Dissertation (PDF):&lt;/strong&gt;
📄&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;Defense recording:&lt;/strong&gt;
🎥&lt;/li&gt;
&lt;/ul&gt;
&lt;h2 id="doctoral-committee"&gt;Doctoral Committee&lt;/h2&gt;
&lt;ul&gt;
&lt;li&gt;Rector Magnificus, &lt;em&gt;Chairman&lt;/em&gt;&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;Prof. dr. K. G. Langendoen&lt;/strong&gt;, Delft University of Technology — &lt;em&gt;promotor&lt;/em&gt;&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;Dr. G. Iosifidis&lt;/strong&gt;, Delft University of Technology — &lt;em&gt;copromotor&lt;/em&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;h3 id="independent-members"&gt;Independent Members&lt;/h3&gt;
&lt;ul&gt;
&lt;li&gt;&lt;strong&gt;Prof. dr. G. Dán&lt;/strong&gt;, KTH Royal Institute of Technology, Sweden&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;Prof. dr. P. Grosso&lt;/strong&gt;, University of Amsterdam&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;Prof. dr. M. T. J. Spaan&lt;/strong&gt;, Delft University of Technology&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;Prof. dr. T. Spyropoulos&lt;/strong&gt;, Technical University of Crete, Greece&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;Dr. T. A. L. van Erven&lt;/strong&gt;, University of Amsterdam&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;I am deeply grateful to my promotors, the members of the committee, my colleagues, and my family
for their support throughout this journey.&lt;/p&gt;</description></item></channel></rss>