BEGIN:VCALENDAR VERSION:2.0 PRODID:-//Date iCal//NONSGML kigkonsult.se iCalcreator 2.20.2// METHOD:PUBLISH X-WR-CALNAME;VALUE=TEXT:Eventi DIAG BEGIN:VTIMEZONE TZID:Europe/Paris BEGIN:STANDARD DTSTART:20171029T030000 TZOFFSETFROM:+0200 TZOFFSETTO:+0100 TZNAME:CET END:STANDARD BEGIN:DAYLIGHT DTSTART:20170326T020000 TZOFFSETFROM:+0100 TZOFFSETTO:+0200 RDATE:20180325T020000 TZNAME:CEST END:DAYLIGHT END:VTIMEZONE BEGIN:VEVENT UID:calendar.12428.field_data.0@www.ugov-ricerca.uniroma1.it DTSTAMP:20260408T063213Z CREATED:20170801T175937Z DESCRIPTION:Speaker: Tommaso ColomboTitle: Recurrent Neural Networks: why d o LSTM networks perform so well in time series prediction?(Joint work with : Alberto De Santis\, Stefano Lucidi)Abstract:Long Short-Term Memory (LSTM ) networks are a broadly-used and very well-known variant of Recurrent Neu ral Networks\, heavily employed\, e.g.\, in time series forecasting and na tural language processing. Deep learning has been revolutionary in many fi elds for the last two decades (e.g.\, image recognition\, natural language processing\, ...) thanks to the continuous increase in computing speed an d capacity. To train deep learning machines\, one needs to solve complex\, highly nonlinear and large-scale optimization problems\, but this notwith standing few authors studied the theoretical properties of such problems. Based on their structure\, Recurrent Neural Networks are naturally the mos t suitable to solve a time series forecasting problem\, but their training leads to one of the most difficult optimization problems in Deep Learning . This difficulty is strictly tied to the vanishing gradients problem that arises when trying to latch information for a long time period.There exis t two main approaches to overcome these training issues: a structural one (e.g.\, LSTM and other memory-based networks) and an algorithmic one (e.g. \, gradient truncation\, …). The most effective\, recent approaches usuall y employ a combination of the two\, i.e. a structure tailored to the probl em and an optimization algorithm tailored to the structure at hand.------- -------------------------------Speaker: Ludovica MaccarroneTitle: A new gr ey-box approach to solve the workforce scheduling problem in complex manuf acturing and logistic contexts(Joint work with Stefano Lucidi)Abstract:We present a new approach to solve the workforce scheduling problem in comple x applicative contexts such as manufacturing and logistic processes.We con sider systems where one or more workloads require to be sequentially proce ssed in different areas by different types of operators exclusively charac terized by their skills. We assume the request of such skills is not fixed and may be varied in order to match the time/cost objectives of the organ ization. Furthermore\, due to the complexity of the considered processes\, we suppose it is not possible to derive an analytic expression linking th e number of resources of different types working on an activity to the tim e to complete it. For this reason\, a set of ad hoc simulators can be empl oyed and their outputs are parameters of our formulation.Typical issues ar ising in workforce management applications are related to the need of mini mizing the labor cost while meeting deadlines and industrial plans. These resource/time trade-offs are even more complex under our assumptions due t o the presence of simulators which natively split the problem into two seq uential sub-problems. Our strategy addresses these difficulties through a decomposition approach which allows to model the problem as a grey-box opt imization problem combining a new scheduling formulation with the simulati on of some complex manufacturing and logistic processes.  DTSTART;TZID=Europe/Paris:20170831T110000 DTEND;TZID=Europe/Paris:20170831T120000 LAST-MODIFIED:20200515T073055Z LOCATION:Aula A2 - DIAG SUMMARY:MORE@DIAG - Speakers: Tommaso Colombo\; Ludovica Maccarrone - Tomm aso Colombo\, Ludovica Maccarrone URL;TYPE=URI:http://www.ugov-ricerca.uniroma1.it/node/12428 END:VEVENT END:VCALENDAR