Valentyn Melnychuk 🇺🇦
@valik-melnychuk.bsky.social
PhD Candidate at Institute of AI in Management, LMU Munich
causal machine learning, causal inference
causal machine learning, causal inference
🧵 4/4
🔗Link to the UA bank:
send.monobank.ua/jar/AK5rgV1iAT
🔗Link to PayPal (if the first option doesn’t work):
paypal.me/vmelnychuk
🔗Link to the UA bank:
send.monobank.ua/jar/AK5rgV1iAT
🔗Link to PayPal (if the first option doesn’t work):
paypal.me/vmelnychuk
Безпечний переказ коштів
Надсилайте безкоштовно та безпечно кошти
send.monobank.ua
April 16, 2025 at 11:22 AM
🧵 4/4
🔗Link to the UA bank:
send.monobank.ua/jar/AK5rgV1iAT
🔗Link to PayPal (if the first option doesn’t work):
paypal.me/vmelnychuk
🔗Link to the UA bank:
send.monobank.ua/jar/AK5rgV1iAT
🔗Link to PayPal (if the first option doesn’t work):
paypal.me/vmelnychuk
🧵 3/3
Thus, his sister, Julia, and I started a fundraiser for a DJI Mavic 3T drone equipped with a thermal camera. This is perfect for night-time intelligence and monitoring.
📍Goal of the first fundraiser (1/2): 100,000 UAH (2,100 euros)
📍Total goal: 255,000 UAH (5,400 euros)
💙💛
Thus, his sister, Julia, and I started a fundraiser for a DJI Mavic 3T drone equipped with a thermal camera. This is perfect for night-time intelligence and monitoring.
📍Goal of the first fundraiser (1/2): 100,000 UAH (2,100 euros)
📍Total goal: 255,000 UAH (5,400 euros)
💙💛
April 16, 2025 at 11:22 AM
🧵 3/3
Thus, his sister, Julia, and I started a fundraiser for a DJI Mavic 3T drone equipped with a thermal camera. This is perfect for night-time intelligence and monitoring.
📍Goal of the first fundraiser (1/2): 100,000 UAH (2,100 euros)
📍Total goal: 255,000 UAH (5,400 euros)
💙💛
Thus, his sister, Julia, and I started a fundraiser for a DJI Mavic 3T drone equipped with a thermal camera. This is perfect for night-time intelligence and monitoring.
📍Goal of the first fundraiser (1/2): 100,000 UAH (2,100 euros)
📍Total goal: 255,000 UAH (5,400 euros)
💙💛
🧵 2/3
My cousin is now on the front line in Ukraine, where every day, together with his companions, he risks his life for the sake of our (and European) freedom. The equipment they use tends to break down, and the situation on the field changes quickly, so a night drone is urgently needed.
My cousin is now on the front line in Ukraine, where every day, together with his companions, he risks his life for the sake of our (and European) freedom. The equipment they use tends to break down, and the situation on the field changes quickly, so a night drone is urgently needed.
April 16, 2025 at 11:20 AM
🧵 2/3
My cousin is now on the front line in Ukraine, where every day, together with his companions, he risks his life for the sake of our (and European) freedom. The equipment they use tends to break down, and the situation on the field changes quickly, so a night drone is urgently needed.
My cousin is now on the front line in Ukraine, where every day, together with his companions, he risks his life for the sake of our (and European) freedom. The equipment they use tends to break down, and the situation on the field changes quickly, so a night drone is urgently needed.
That performance wasn’t live though 🙃
January 28, 2025 at 6:30 PM
That performance wasn’t live though 🙃
5/n 📄 See the full paper: arxiv.org/abs/2411.03387
Quantifying Aleatoric Uncertainty of the Treatment Effect: A Novel Orthogonal Learner
Estimating causal quantities from observational data is crucial for understanding the safety and effectiveness of medical treatments. However, to make reliable inferences, medical practitioners requir...
arxiv.org
December 13, 2024 at 6:50 PM
5/n 📄 See the full paper: arxiv.org/abs/2411.03387
4/n Many thanks to @zuseschoolrelai.bsky.social and @munichcenterml.bsky.social the financial support.
December 13, 2024 at 6:50 PM
4/n Many thanks to @zuseschoolrelai.bsky.social and @munichcenterml.bsky.social the financial support.
3/n Hence, we target at the partial identification bounds, given by Makorov bounds.
👉 We develop a novel Neyman-orthogonal learner to learn covariate-conditional Makarov bounds, namely AU-learner.
👉 We develop a novel Neyman-orthogonal learner to learn covariate-conditional Makarov bounds, namely AU-learner.
December 13, 2024 at 6:49 PM
3/n Hence, we target at the partial identification bounds, given by Makorov bounds.
👉 We develop a novel Neyman-orthogonal learner to learn covariate-conditional Makarov bounds, namely AU-learner.
👉 We develop a novel Neyman-orthogonal learner to learn covariate-conditional Makarov bounds, namely AU-learner.
2/n We study the aleatoric uncertainty of the treatment effect at the individualized level: the covariate-conditional distribution of the treatment effect.
Surprisingly, this is not identifiable even in RCTs.
Surprisingly, this is not identifiable even in RCTs.
December 13, 2024 at 6:48 PM
2/n We study the aleatoric uncertainty of the treatment effect at the individualized level: the covariate-conditional distribution of the treatment effect.
Surprisingly, this is not identifiable even in RCTs.
Surprisingly, this is not identifiable even in RCTs.