<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"><channel><description>PhD Candidate at Institute of AI in Management, LMU Munich&#xA;causal machine learning, causal inference</description><link>https://bsky.app/profile/valik-melnychuk.bsky.social</link><title>@valik-melnychuk.bsky.social - Valentyn Melnychuk 🇺🇦</title><item><link>https://bsky.app/profile/valik-melnychuk.bsky.social/post/3mpxtqrfr4k2b</link><description>[1/4] 🇰🇷 Excited to present our new paper “Frequentist Consistency of Prior-Data Fitted Networks for Causal Inference” at ICML 2026 in Seoul.&#xA;&#xA;The paper asks a simple question: when PFNs are used for causal inference, do their uncertainty estimates actually agree with classical frequentist inference?</description><pubDate>06 Jul 2026 10:06 +0000</pubDate><guid isPermaLink="false">at://did:plc:hlmvpsdyr2ymzfvwfaocyniq/app.bsky.feed.post/3mpxtqrfr4k2b</guid></item><item><link>https://bsky.app/profile/valik-melnychuk.bsky.social/post/3mkz6js5cj22h</link><description>[1/5] 🇲🇦 Today I’m presenting our paper “Orthogonal Representation Learning for Estimating Causal Quantities” at AISTATS 2026 in Tangier, Morocco; and I’m very happy that it was accepted as an oral.</description><pubDate>04 May 2026 07:46 +0000</pubDate><guid isPermaLink="false">at://did:plc:hlmvpsdyr2ymzfvwfaocyniq/app.bsky.feed.post/3mkz6js5cj22h</guid></item><item><link>https://bsky.app/profile/valik-melnychuk.bsky.social/post/3mk6mqiiur22p</link><description>[1/5] 🇧🇷 Looking forward to presenting our new paper, “Overlap-Adaptive Regularization for Conditional Average Treatment Effect Estimation,” at ICLR 2026 in Rio de Janeiro.&#xA;The paper tackles a core challenge in causal ML: estimating treatment effects when overlap is low.</description><pubDate>23 Apr 2026 18:18 +0000</pubDate><guid isPermaLink="false">at://did:plc:hlmvpsdyr2ymzfvwfaocyniq/app.bsky.feed.post/3mk6mqiiur22p</guid></item><item><link>https://bsky.app/profile/valik-melnychuk.bsky.social/post/3mk6dplm3g22f</link><description>[1/4] 🇧🇷 Excited to share our new paper, “GDR-learners: Orthogonal Learning of Generative Models for Potential Outcomes,” which I’ll be presenting at ICLR 2026.&#xA;The goal: better estimate potential outcome distributions from observational data — not just averages, but the full uncertainty.</description><pubDate>23 Apr 2026 15:36 +0000</pubDate><guid isPermaLink="false">at://did:plc:hlmvpsdyr2ymzfvwfaocyniq/app.bsky.feed.post/3mk6dplm3g22f</guid></item><item><link>https://bsky.app/profile/valik-melnychuk.bsky.social/post/3lmwhtqzybs2n</link><description>🧵 1/3&#xA;&#xA;Dear friends, I ask everyone again for help! &#xA;&#xA;You know the situation in Ukraine is tough at the moment: After the deadly attack on Sumy in Ukraine last Sunday I felt speechless. &#xA;&#xA;Yet there is a way to help and even a small contribution might eventually save someone’s life.</description><pubDate>16 Apr 2025 11:18 +0000</pubDate><guid isPermaLink="false">at://did:plc:hlmvpsdyr2ymzfvwfaocyniq/app.bsky.feed.post/3lmwhtqzybs2n</guid></item><item><link>https://bsky.app/profile/valik-melnychuk.bsky.social/post/3ld7gzmkh722b</link><description>1/n  Excited to present our new work, done at the van der Schaar Lab at the Cambridge University together with Mihaela van der Schaar and Stefan Feuerriegel. &#xA;📍7008 western ballroom, 11:00-14:00&#xA;#NeurIPS2024</description><pubDate>13 Dec 2024 18:46 +0000</pubDate><guid isPermaLink="false">at://did:plc:hlmvpsdyr2ymzfvwfaocyniq/app.bsky.feed.post/3ld7gzmkh722b</guid></item><item><link>https://bsky.app/profile/valik-melnychuk.bsky.social/post/3ld2ih4bcfs2v</link><description>✈️ Attending #NeurlPS2024. If you fancy grabbing a coffee and chatting about causal machine learning, drop me a message</description><pubDate>11 Dec 2024 19:28 +0000</pubDate><guid isPermaLink="false">at://did:plc:hlmvpsdyr2ymzfvwfaocyniq/app.bsky.feed.post/3ld2ih4bcfs2v</guid></item></channel></rss>