<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40"><head><meta http-equiv=Content-Type content="text/html; charset=us-ascii"><meta name=Generator content="Microsoft Word 15 (filtered medium)"><style><!--
/* Font Definitions */
@font-face
{font-family:"Cambria Math";
panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
{font-family:Calibri;
panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0in;
margin-bottom:.0001pt;
font-size:11.0pt;
font-family:"Calibri",sans-serif;}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:#0563C1;
text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
{mso-style-priority:99;
color:#954F72;
text-decoration:underline;}
p
{mso-style-priority:99;
mso-margin-top-alt:auto;
margin-right:0in;
mso-margin-bottom-alt:auto;
margin-left:0in;
font-size:12.0pt;
font-family:"Times New Roman",serif;}
pre
{mso-style-priority:99;
mso-style-link:"HTML Preformatted Char";
margin:0in;
margin-bottom:.0001pt;
font-size:10.0pt;
font-family:"Courier New";}
span.EmailStyle17
{mso-style-type:personal-compose;
font-family:"Calibri",sans-serif;
color:windowtext;}
span.HTMLPreformattedChar
{mso-style-name:"HTML Preformatted Char";
mso-style-priority:99;
mso-style-link:"HTML Preformatted";
font-family:"Courier New";}
.MsoChpDefault
{mso-style-type:export-only;
font-family:"Calibri",sans-serif;}
@page WordSection1
{size:8.5in 11.0in;
margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]--></head><body lang=EN-US link="#0563C1" vlink="#954F72"><div class=WordSection1><p class=MsoNormal><span style='font-size:12.0pt;font-family:"Times New Roman",serif'>INSTITUTE FOR ADVANCED STUDY<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:12.0pt;font-family:"Times New Roman",serif'>School of Mathematics<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:12.0pt;font-family:"Times New Roman",serif'>Princeton, NJ 08540<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:12.0pt;font-family:"Times New Roman",serif'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:12.0pt;font-family:"Times New Roman",serif'>Mathematics Seminars</span><span style='font-size:12.0pt;font-family:"Times New Roman",serif'><o:p></o:p></span></p><pre><span style='font-size:12.0pt;font-family:"Times New Roman",serif'>Thursday, April 12<o:p></o:p></span></pre><pre><span style='font-size:12.0pt;font-family:"Times New Roman",serif'><o:p> </o:p></span></pre><pre><span style='font-size:12.0pt;font-family:"Times New Roman",serif'><o:p> </o:p></span></pre><pre style='margin-bottom:3.0pt'><b><span style='font-size:12.0pt;font-family:"Times New Roman",serif'>Seminar on Theoretical Machine Learning<o:p></o:p></span></b></pre><pre><span style='font-size:12.0pt;font-family:"Times New Roman",serif'>Topic: Stability and Generalization in Adaptive Data Analysis<o:p></o:p></span></pre><pre><span style='font-size:12.0pt;font-family:"Times New Roman",serif'>Speaker: Vitaly Feldman, Google Brain<o:p></o:p></span></pre><pre><span style='font-size:12.0pt;font-family:"Times New Roman",serif'>Time/Room: 12:15pm - 1:45pm/White-Levy Room<o:p></o:p></span></pre><pre><span style='font-size:12.0pt;font-family:"Times New Roman",serif'>Abstract Link: <a href="http://www.math.ias.edu/seminars/abstract?event=135273">http://www.math.ias.edu/seminars/abstract?event=135273</a><o:p></o:p></span></pre><p>Datasets are often used multiple times with each successive analysis depending on the outcomes of previous analyses on the same dataset. Standard techniques for ensuring generalization and statistical validity do not account for this adaptive dependence. In this talk I will overview a recent line of work on adaptive data analysis that studies the problem of answering a sequence of "queries" about the data distribution where each query may depend arbitrarily on answers to previous queries. I'll then focus on a new notion of algorithmic stability with the important property that it "composes" well when data is reused. I'll demonstrate that this notion of stability implies that the algorithm reveals little information about its input and, in particular, cannot lead to overfitting. Finally, I'll describe simple algorithms based on this notion that can answer statistical queries about the dataset with substantially improved accuracy guarantees for low-variance queries.<br> <br>Based in part on a joint work with Thomas Steinke.<o:p></o:p></p></div></body></html>