<div id="__MailbirdStyleContent" style="font-size: 10pt;font-family: Arial;color: #000000;text-align: left" dir="ltr">Perjantaina taas kollokviopuhe!<div><br></div><div>~ Antero Voutilainen<br><div class="mb_sig"></div><div class="history_container"><p style="color: margin-top: 10px;">------ Forwarded Message --------<br>From: Tero Tapio Heikkilä <tero.t.heikkila@jyu.fi><br>Date: 11.3.2024 8.33.36 <br>Subject: Tommi Kärkkäinen: About machine learning methods: A quest for simplicity and transparency, in the physics colloquium on Friday<span style="font-family: Arial, Helvetica, sans-serif;font-size: 10pt">Welcome to the University of Jyväskylä physics colloquium. </span><br></p><div style="font-family:Arial,Helvetica,sans-serif">
<p> On Friday 15th March at 10 am in FYS1 and <a href="https://jyufi.zoom.us/j/66703175507">Zoom</a> (passcode
890524):</p>
<p><b>Tommi Kärkkäinen</b>, University of Jyväskylä</p>
<p><i>About machine learning methods: A quest for simplicity and
transparency</i></p>
<p><span style="caret-color: rgb(0, 0, 0); color: rgb(0, 0, 0); font-style: normal; font-variant-caps: normal; font-weight: 400; letter-spacing: normal; orphans: auto; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: auto; word-spacing: 0px; -webkit-text-stroke-width: 0px; text-decoration: none;"><span style="font-size: 10pt">Types of data-driven machine learning consist of
supervised, unsupervised, and reinforcement learning methods.
Here, I will discuss unsupervised and supervised methods,
which have provided excellent results in various applications.
The amazing results, however, might be difficult to reproduce.
In fact, there are many aspects in relation to especially
currently popular deep learning techniques which makes their
use complicated and nontransparent. Surprisingly, by imitating
classical linear regression, simple but yet powerful machine
learning approaches can be derived. In this presentation, I
will present and exemplify these techniques and some of their
background in the context of monolayer-protected clusters.</span></span></p>
<p>Welcome!</p>
<p> Coffee will be served in the lobby. Although on-site attendance
is the preferred option, you can also join via Zoom. Please do not
send chat messages in Zoom during the talk, except if you need to
tell about a muted speaker. <br>
</p>
<p> Confirmed colloquium talks in the Spring 2024 (you are welcome
to suggest more - there are still open slots):</p>
<p>5.4. Heikki Mäntysaari (JYU): <i>TBA<br>
</i>12.4. Sebastian Bergeret (San Sebastian): <i>Superconducting
spintronics</i></p>
<p>Tero & Iain<br>
</p>
<p></p>
<p></p>
<pre class="moz-signature" cols="72">--
Tero Heikkilä
Professor, Department of Physics
University of Jyväskylä</pre>
</div></div>
</div></div><img class="mailbird" width="1" height="1" style="border:0; width:1; height:1; display: none;" src="https://tracking.getmailbird.com/OpenTrackingPixel/?messageId=Mailbird-76a3f18c-9dae-4ccf-b5cd-5fb9017f3648@jyfk.fi&senderHash=D3DD145366F33B8ED5A4EFBE2EEFC1E313854D8BD84B417068A92F36BEDFC2AF&recipientHash=558237956A365CC02C7581D7A1C76DE20986D31F6736097F936EEB8E4C823B68&internalId=dabeb082-9ef1-4e54-b440-66fccaa019f4" alt="65301ef2-cf3a-47fb-b75e-6fa13d159792">