top of page

Emille E. O. Ishida, PhD

I am a Brazilian physicist based in France and  working in Astronomy and Cosmology since I can remember.  

My research is focused on machine learning applications to astronomy and in the development of sustainable interdisciplinary scientific environments.

I am co-founder of the Cosmostatistics Initiative (COIN), 

the Fink broker and the SNAD team. 

Main scientific activities

An adaptable LSST community broker based on machine learning
The Cosmostatistics Initiative
SuperNova Anomaly Detection


Screenshot from 2024-06-14 13-19-02.png

ELEPHANT: ExtragaLactic alErt Pipeline for Hostless AstroNomical Transients

Pessi et al., 2024

We present the ExtragaLactic alErt Pipeline for Hostless AstroNomical Transients (ELEPHANT), a framework for filtering hostless transients in astronomical data streams. We used Fink to access all the ZTF alerts produced between January/2022 and December/2023, selecting only those associated with extragalactic transients. We then processed the associated stamps using a sequence of image analysis techniques to retrieve hostless candidates. We find that less than 2% of all analyzed transients are potentially hostless. Among them, approximately 10% have a spectroscopic class reported on TNS, with Type Ia supernova being the most common class, followed by SLSN. Among the hostless candidates retrieved by our pipeline, there was SN 2018ibb, which has been proposed to be a PISN candidate; and SN 2022ann, one of only five known SNe Icn. When no class is reported on TNS, the dominant classes are QSO and SN candidates, the former obtained from SIMBAD and the latter inferred using the Fink ML classifier. ELEPHANT represents an effective strategy to filter extragalactic events within large and complex astronomical alert streams. There are many applications for which this pipeline will be useful, ranging from transient selection for follow-up to studies of transient environments.

Screenshot from 2024-04-18 11-01-22.png

 Transient Classifiers for Fink: Benchmarks for LSST

Fraga et al., 2024


The upcoming Legacy Survey of Space and Time (LSST) at the Vera Rubin Observatory is expected to detect a few million transients per night, which will generate a live alert stream during the entire 10 years of the survey. This will be distributed via community brokers whose task is to select subsets of the stream and direct them to scientific communities. Given the volume and complexity of data, machine learning (ML) algorithms will be paramount for this task. We present the infrastructure tests and classification methods developed within the {\sc Fink} broker in preparation for LSST. This work aims to provide detailed information regarding the underlying assumptions, and methods, behind each classifier, enabling users to make informed follow-up decisions from {\sc Fink} photometric classifications. Using simulated data from the Extended LSST Astronomical Time-series Classification Challenge (ELAsTiCC), we showcase the performance of binary and multi-class ML classifiers available in {\sc Fink}. These include tree-based classifiers coupled with tailored feature extraction strategies, as well as deep learning algorithms. We introduce the CBPF Alert Transient Search (CATS), a deep learning architecture specifically designed for this task. Results show that {\sc Fink} classifiers are able to handle the extra complexity which is expected from LSST data. CATS achieved 97% accuracy on a multi-class classification while our best performing binary classifier achieve 99% when classifying the Periodic class. ELAsTiCC was an important milestone in preparing {\sc Fink} infrastructure to deal with LSST-like data. Our results demonstrate that {\sc Fink} classifiers are well prepared for the arrival of the new stream; this experience also highlights that transitioning from current infrastructures to Rubin will require significant adaptation of currently available tools.

bottom of page