מכתב לסטודנטים

 

מערכת שעות לסמסטר קיץ תשע"ט

 

החל מה 22.7.19

בין התאריכים ה-25.8.19-31.8.19 האוניברסיטה סגורה לרגל חופשה מרוכזת

 

 

ייתכנו שינויים במערכת השעות – פתיחת הקורסים והמעבדות מותנה במספר מינימלי של נרשמים

 

 

 

 

0509.1829         פיזיקה (2)                    ש' 4 + ת' 2

 

קבוצת שיעור

 

יום

שעות

תאריכים

20

הרצאה

א'

09:30-13:30

28/7,4/8,18/8

 

 

ג'

09:30-13:30

23/7,30/7,6/8,13/8,20/8

 

 

ה'

09:30-12:30

25/7,1/8,8/8,15/8,22/8

 

 

 

 

 

21

תרגול

א'

14:00-16:00

28/7,4/8,18/8

 

 

ג'

14:00-16:00

23/7,30/7,6/8,13/8,20/8

 

 

ה'

13:00-15:00                                                 

25/7,1/8,8/8,15/8,22/8

 

 

 

 

 

22

תרגול

א'

14:00-16:00

28/7,4/8,18/8

 

 

ג'

14:00-16:00

23/7,30/7,6/8,13/8,20/8

 

 

ה'

13:00-15:00                                                 

25/7,1/8,8/8,15/8,22/8

 

 

 

 

 

 

 

 

 

 

 

11.8 צום ט' באב לא יתקיימו לימודים

 

 

 

 

 

 

 

 

בחינה מועד א' – 3.9.2019

בחינה מועד ב' – 6.10.2019

 

 

 

 

 

 

0509.2801           מבוא להסתברות וסטטיסטיקה            ש' 3 + ת' 1

 

קבוצת שיעור

 

יום

שעות

תאריכים

10

הרצאה

ב'

09:00-12:00                                           

22/7,29/7,5/8,12/8,19/8

 

 

ג'

09:00-12:00                                           

23/7,30/7,6/8,13/8,20/8

 

 

ה'

09:00-11:00                                           

25/7,1/8,8/8,15/8,22/8

 

 

 

 

 

11

תרגיל

ב'

13:00-14:30

22/7,29/7,5/8,12/8,19/8

 

 

ה'

11:00-12:00

25/7,1/8,8/8,15/8,22/8

 

 

 

 

 

 

 

בחינה מועד א' – 3.9.2019

בחינה מועד ב' – 6.10.2019

 

 

 

 

 

 

 

0509.2804            אנליזה נומרית                  ש' 3 + ת' 1

 

קבוצת שיעור

 

יום

שעות

תאריכים

20

הרצאה

א'

09:00-12:00

28/7,4/8,18/8

 

 

ג'

09:00-12:00

23/7,30/7,6/8,13/8,20/8

 

 

ד'

09:00-12:00

24/7,31/7,7/8,14/8,21/8

 

 

 

 

 

21

תרגול

א'

12:30-13:30

28/7,4/8,18/8

 

 

ד'

13:00-15:00

24/7,31/7,7/8,14/8,21/8

 

 

 

 

 

 

11.8 צום ט' באב לא יתקיימו לימודים

 

 

בחינה מועד א' – 3.9.2019

בחינה מועד ב' – 6.10.2019

 

 

 

0512.3592         אלקטרוניקה – מעבדה (2)          מ' 4

 

קבוצת שיעור

יום

שעות

תאריכים

07

ב'

09:00-13:00

22/7,29/7,5/8,12/8,19/8,2/9,9/9

 

ה'

09:00-13:00

25/7,1/8,8/8,15/8,22/8,5/9

 

 

 

 

08

ב'

14:00-18:00

22/7,29/7,5/8,12/8,19/8,2/9,9/9

 

ה'

14:00-18:00

25/7,1/8,8/8,15/8,22/8,5/9

 

 

 

 

 

 

0512.3593              אלקטרוניקה-מעבדה (3) מ' 4    

                                                                                               

קבוצת שיעור

יום

שעות

תאריכים

07

ג'

09:00-13:00

23/7,30/7,6/8,13/8,20/8,3/9,10/9

 

ד'

09:00-13:00

24/7,31/7,7/8,14/8,21/8,4/9

 

 

*לתשומת ליבכם ב-20.8.19 המעבדה תתקיים בין השעות 8:00-13:00, בחדר 008-תוכנה. (ולא מ-9:00 עד 13:00 בחדר 306 וולפסון כמו בדרך כלל)

 

 

  

                0542.4092               זרימה ומעבר חום – מעבדה מ' 3   (27 מקומות בקבוצה)

 

                                               

 

 

קבוצת שיעור

יום

שעות

תאריכים

05

ב'

09:00-12:00

22/7,29/7,5/8,12/8,19/8,

 

ד'

09:00-12:00

24/7,31/07,07/08,14/08,21/8

 

 

 

בתאריכים 31/7 ו 07/08 יבוצעו שתי מעבדות במתכונת של שתי משמרות :

משמרת בוקר 9-12 משמרת צהריים 12-15.

 

בתאריכים19/8 ו 21/8 יתקיימו השלמות +החזרת דוחות .

EE Seminar: on sample compression

15 באפריל 2019, 15:00 
חדר 011, בניין כיתות-חשמל  

(The talk will be given in English)

 

Speaker:     Prof. Aryeh Kontorovich
                     Department of Computer Science, BGU

 

Monday, April 15th, 2019
15:00 - 16:00

Room 011, Kitot Bldg., Faculty of Engineering

 

on sample compression
 

Abstract

Sample compression is a natural and elegant learning framework, which allows for storage and runtime savings as well as sharp generalization bounds. In this talk, I will survey a few recent collaborations that touch upon various aspects of sample compression. Central among these is the development of a new algorithm for learning in arbitrary metric spaces based on a margin-regularized 1-nearest neighbor, which we call OptiNet. The latter is strongly universally Bayes-consistent in all essentially-separable metric probability spaces. OptiNet is the first learning algorithm to enjoy this property; by comparison, k-NN and its variants are not Bayes-consistent, except under additional structural assumptions, such as an inner product, a norm, finite doubling dimension, or a Besicovitch-type property. I will then talk about sample compression in the context of regression, extensions to non-uniform margins, and, time permitting, eneralization lower bounds.
Based on joint work with: Lee-Ad Gottlieb, Steve Hanneke, Sivan Sabato, Menachem Sadigurschi, Roi Weiss

Short Bio: Aryeh Kontorovich received his undergraduate degree in mathematics with a certificate in applied mathematics from Princeton University in 2001. His M.Sc. and Ph.D. are from Carnegie Mellon University, where he graduated in 2007. After a postdoctoral fellowship at the Weizmann Institute of Science, he joined the Computer Science department at Ben-Gurion University of the Negev in 2009, where he is currently an associate professor. His research interests are mainly in machine learning, with a focus on probability, statistics, Markov chains and metric spaces.

EE Seminar: Fused-VIVT: Reducing Power by Coalescing the Cache Directory and TLB

10 באפריל 2019, 15:30 
חדר 011, בניין כיתות-חשמל  

 

Speaker: Eyal Naor

M.Sc. student under the supervision of Prof. Shlomo Weiss

 

Wednesday, April 10th, 2019 at 15:30

Room 011, Kitot Bldg., Faculty of Engineering

 

Fused-VIVT: Reducing Power by Coalescing the Cache Directory and TLB

 

Abstract

We propose a novel Fused-VIVT that merges the directories of the cache and TLB. The main benefit of this design is substantial power reduction in accessing cache directories, without readily affecting performance, area requirements and latency. When used in conjunction with a way-predictor, the energy consumption can be reduced by 90% compared to the traditional cache directory and TLB access scheme. The present research provides an evaluation of this proposal in these important aspects: energy consumption, area requirements, latency and performance. We also provide a comparison and analysis of a “traditional” micro-architecture consisting of a TLB and a physically tagged cache, with that of a Fused-VIVT. We also address various difficulties that could arise with implementing such a cache, including synonyms and coherency. 

 

EE Seminar: Formation Control in Networked Dynamical Systems

29 באפריל 2019, 15:00 
חדר 011, בניין כיתות-חשמל  

(The talk will be given in English)

 

Speaker:     Anoop Jain
                     Faculty of Aerospace Engineering, Technion

 

Monday, April 29th, 2019
15:00 - 16:00

Room 011, Kitot Bldg., Faculty of Engineering

 

Formation Control in Networked Dynamical Systems
 

Abstract

Networked dynamical systems possess improved robustness properties over single-agent systems, and hence, are desirable for various engineering applications ranging from searching and sensing, tracking and monitoring, surveillance and patrolling to cooperative space missions. To accomplish such cooperative tasks autonomously, the agents are required to coordinate with each other depending on their sensing capabilities, which is represented by a graph (usually referred to as communication network).

This talk will focus on achieving a collective motion in multi-agent systems where the agents are either in a synchronized or a balanced formation. Such formations are characterized by the motion of the collective centroid of the group of agents and serve as motion primitives to achieve general tracking patterns. In the existing literature, the problems of achieving synchronization and balancing have been studied under the assumption that the agents are coupled through controller gains that are identical or homogeneous. In this talk, I plan to discuss a practical scenario where the gains may vary nominally due to minor implementation errors or drastically due to major faults or errors, thus leading to heterogeneity in controller gains. Robustness and fault tolerant properties of these formations against the failure of controllers will also be presented. Furthermore, stabilization of more general formation patterns, such that the trajectories of the agents do not transgress a given workspace, will also be discussed.

School of Mechanical Engineering Dr. Raya Sorkin

29 באפריל 2019, 14:00 - 15:00 
בניין וולפסון חדר 206  
0
School of Mechanical Engineering Dr. Raya Sorkin

EE Seminar: On the Information Complexity of Learning

08 באפריל 2019, 15:00 
חדר 011, בניין כיתות-חשמל  

(The talk will be given in English)

 

Speaker:     Ido Nachum
                     Department of Mathematics, Technion

 

Monday, April 8th, 2019
15:00 - 16:00

Room 011, Kitot Bldg., Faculty of Engineering

 

On the Information Complexity of Learning
 

Abstract

Under an information theoretic setting, these two notions are indeed equivalent. 
a) Compression implies learning. We will show that learning algorithms that retain a small amount of information from their input sample generalize.
b) Learning implies compression. We will show that under an average-case analysis, every hypothesis class of finite VC dimension (a characterization of learnable classes) has empirical risk minimizers (ERM) that do not reveal a lot of information.

If time permits, we will discuss a worst-case lower bound by presenting a simple concept class for which every empirical risk minimizer (possibly randomized) must reveal a lot of information and also observe connections with well-studied notions such as sample compression schemes, Occam's razor, PAC-Bayes, and differential privacy.

יום עיון בחסות חברת סיסטמטיקס בו גילינו חידושים, טיפים וטריקים בסביבת Matlab

31 מרץ 2019
יום עיון עם חברת סיסתמטיקס

בתאריך ה 24.3.2019 חברת סיסטמטיקס בשיתוף ארגון עמיתי התעשייה והפקולטה להנדסה אוניברסיטת תל אביב קיימה יום עיון בנושא MATLAB היועד לחברי הסגל, סטודנטים לתארים מתקדמים מכל הפקולטות בדגש מיוחד לפקולטה למדעים מדוייקים ולפקולטה להנדסה.

 

מרצה אורחת

יום העיון הועבר ע"י ד"ר לורן שור העובדת בחברת MathWorks  מעל 30 שנה (היא העובדת הראשונה שנשכרה לעבוד בחברה על ידי המנכ"ל). ד"ר שור אחראית על פיתוח פונקציות רבות של MATLAB וה-Toolboxes השונים. כיום היא עוסקת בעיצוב שפת ה-MATLAB העתידית ובכתיבת הבלוג הפופולרי שלה - The Art of MATLAB. במהלך ביקורה הקודם בישראל, לפני כ-4 שנים, הרצתה Dr. Shure בכנסים וסמינרים במוסדות האקדמיים המובילים ברחבי הארץ, בהשתתפות יותר מ- 1000 איש. הרצאותיה מעניינות, עדכניות וזוכות לפופולריות רבה ברחבי העולם. MATLAB הינה תוכנה מתמטית המיועדת לביצוע חישובי, טכניי והנדסי ומיוצרת ע"י חברת MathWorks

 

מה למדנו?

ביום עיון זה למדו כיצד Data Analytics יכול להפוך כמויות עצומות של מידע מורכב לכדי מידע שמאפשר קבלת החלטות מושכלות וביצוע פעולות בהתאם. עוד נלמד כיצד לתכנת בסביבת MATLAB בצורה יעילה יותר, חידושים, טיפים וטריקים בנושאים כמו ניהול זיכרון, טיפוסי מידע חדשיםמידע חדשים, בניית Apps תמיכה בחומרה ועוד.

 

על הקשר עם חברת סיסטמטיקס 

חברת סיסטמטיקס, בין הוותיקות בתעשייה הטכנולוגית (משנת 1979) ומעסיקה למעלה מ 150 עובדים, מקדמת את האפקטיביות והיעילות הארגונית, על ידי שיווק טכנולוגיות מיחשוב חדשניות, ויישומן בתהליכי עבודה מתקדמים. סיסטמטיקס הצטרפה השנה כחברה לארגון עמיתי התעשייה של הפקולטה להנדסה. ארגון עמיתי התעשייה הוקם במטרה לייצר תהליך של דיאלוג מתמשך בין התעשייה לבין הפקולטה להנדסה באוניברסיטת תל-אביב. תכלית הארגון הינה ליזום, לחזק ולבסס קשרים ושיתופי פעולה ארוכי טווח בין האקדמיה לבין התעשייה הטכנולוגית בארץ ובעולם. הארגון פעיל כל השנה בחיזוק הקשרים עם התעשיה, וזאת ניכר, בין היתר, בימי זרקור שבועיים, פרסום של משרות, סיורים בחברות והרצאות אורח עם מיטב המומחים בתעשייה. אנחנו מצפים ונרגשים להמשיך לקדם ולפתח את הקשר הפורה עם חברת סיסטמטיקס.

מוזמנים להכנס ללינקים להמשך מעקב אחר ההרצאות של לורן שור למרצים וחוקרים בתחום ההנדסה ומדעים מדוייקים.

גישה:

Username: Customer

Password: MathWorks

What's New: https://sft.mathworks.com/human.aspx?r=&Arg06=503316027&Arg12=filelist

Data Analytics: https://sft.mathworks.com/human.aspx?r=&Arg06=503117509&Arg12=filelist

School of Mechanical Engineering Prof. Stefan Klumpp - Goettingen

27 במאי 2019, 14:00 - 15:00 
בניין וולפסון חדר 206  
0
School of Mechanical Engineering Prof. Stefan Klumpp - Goettingen
JOINT SCHOOL OF MECHANICAL ENGINEERING SEMINAR
& BIOLOGICAL AND SOFT MATTER PHYSICS SEMINAR
Monday, May 27, 2019 at 14:00
Wolfson Building of Mechanical Engineering, Room 206
Swimming with magnets: motility of magnetotactic bacteria
Prof. Stefan Klumpp
Institute for the Dynamics of Complex Systems,
University of Göttingen, Germany
Magnetotatic bacteria align along magnetic field lines with the help of a linear chain of magnetic organelles called magnetosomes. Magnetosomes contain magnetic nanoparticles and are aligned along a cytoskeletal filament, forming a cellular compass needle. Thus, when these bacteria swim by rotating their flagella, they follow the magnetic field, a behavior known as magnetotaxis. These bacteria provide a nice model system to study the interplay of generic physical interactions with specific biological functionality. In the talk, I will discuss structural aspects of the magnetosome chain, specifically its stability and self-assembly, as well as the patterns of magnetotactic motion, which arise from the coupling between magnetotaxis and aerotaxis, based on our combined theoretical and experimental investigations.

 

School of Mechanical Engineering Michael Narznoy

03 באפריל 2019, 14:00 - 15:00 
בניין וולפסון חדר 206  
0
School of Mechanical Engineering Michael Narznoy

 

 

 

 

School of Mechanical Engineering Seminar
Wednesday, April 3, 2019 at 14:00
Wolfson Building of Mechanical Engineering, Room 206

Fishlike underwater imaging

Michael Nareznoy

MSc Student of Gregory Zilman

 

To perform their missions Autonomous Underwater Vehicles have to detect, to avoid and to identify obstacles.  Mature technologies such as underwater video cameras and sonars are used currently for these purposes. However, in low light and turbid water the recognizing a plane wall, a corner, or a finite size body including planes may be difficult.  

 Blind Mexican Tetra fish, are capable to detect, to avoid and even to identify obstacles without vision or sonar. This phenomenon is known in the scientific literature by its figurative name “distant touch” or as hydrodynamic imaging.  Distant touch is based on the alternations of the hydrodynamic field induced by a fish moving in the presence of other bodies and on sensing these alternations by the velocity and pressure sensors of the fish’s lateral line.

Mathematical modelling of the distant touch and understanding how distant and robust it can be is the aim of our work.  Because of a  huge diversity of obstacles shapes, only the very three representative examples are considered: an oblique approach a body-detector to a plane wall, motion of a detector in a corner, and passing by a detector of another body.  A remarkable feature of the lateral line is that the density of the velocity and pressure sensors of a Mexican Tetra reaches the maximum on its forward part (head). For most of smooth bodies, the pressure on the forward (bow) part of them can be modeled as for an incompressible, inviscid and irrotational flow.

Within the framework of potential flow, a multibody problem is studied,  where a simple layer of sources with unknown density replaces the surface of each body.  Using the boundary condition of impermeability the Fredholm second type integral equation is derived.   This integral equation is solved numerically using collocations combined with the panel method. Once the simple layer density is known, it becomes possible to calculate the pressure and the tangential velocity on the body-detector using the Bernoulli integral.

Systematic numerical calculations manifest  two main features of the pressure distribution on the surface of the body-detector when it approaches to an obstacle: i) a sharp rise of pressure on its bow; ii) an essential shift of the stagnation point. While the pressure is proportional to the square of the detector velocity, the shift of the stagnation point is a purely kinematic characteristic.  Notable, Mexican Tetra initiate a turning maneuver from an obstacle also regardless on its speed.  

Finally, possible geometrical configurations of a body-detector and the associated detection distance for three considered examples is discussed.   

School of Mechanical Engineering dana pousty and Ortal Shelah

29 במאי 2019, 14:00 - 15:00 
בניין וולפסון חדר 206  
0
School of Mechanical Engineering dana pousty and Ortal Shelah

 

 

 

 

School of Mechanical Engineering Seminar
Wednesday, May 29, 2019 at 14:00
Wolfson Building of Mechanical Engineering, Room 206

 

Effect of Wavelength and Intensity on E. coli Inactivation Kinetics

Abstract for IUVA Conference

 

Dana Pousty

M.Sc. student of  Prof Hadas Mamane and Ron Hofmann

 

 

1School of Mechanical Engineering, Environmental Engineering Program, Tel Aviv University, Tel Aviv 69978, Israel

2Department of Civil Engineering, University of Toronto, 35 St. George Street, Toronto, Ontario, Canada M5S 1A4

 

UV disinfection efficiency depends on the UV-dose, which is defined as the product of average incident irradiance (fluence rate) and exposure time, corrected by water absorbance, petri factor and reflection factor. It has been reported that the same time-dose reciprocity may not apply to microorganisms when exposed to different light intensities. Sommer et al. 1996, reported that E.coli showed higher UV inactivation when applying a high UV intensity over short exposure times. UV light-emitting diodes (LEDs) are considered as alternatives to UV mercury lamps in water treatment. These UV sources allow flexible design of reactors, and enables tuning the wavelength; require no lamp warm-up time; can be operated by intermittent flow and at ambient temperature to avoid fouling; and can use a DC voltage.

However, LED technology is currently limited to low power input and radiant flux (output). Such low irradiances might affect the expected time-dose reciprocity as described above. The goal of this research was to examine the time-dose reciprocity on E. coli using UV-LEDs at different wavelengths and irradiances, and compare the results to the time dose reciprocity in low-pressure (LP) lamps.

 

The following results were observed: (1) LP - high average intensity over long exposure times result in higher inactivation. (2) LED at 265 nm - no significant differences were observed between the low average intensity and the high average intensity. (3) LED at 275 nm - minor differences were observed between the low average intensity and the high average intensity. (3) LED at 295 and 285 nm - significant differences were observed - low average intensity over long exposure times resulted in higher inactivation.

 

The literature does not provide any explanation for the inactivation kinetic differences between the wavelengths and the intensities. Ongoing research is being conducted to examine whether metabolic mechanisms are affected differently by the different wavelengths and intensities.

עמודים

אוניברסיטת תל אביב עושה כל מאמץ לכבד זכויות יוצרים. אם בבעלותך זכויות יוצרים בתכנים שנמצאים פה ו/או השימוש שנעשה בתכנים אלה לדעתך מפר זכויות
שנעשה בתכנים אלה לדעתך מפר זכויות נא לפנות בהקדם לכתובת שכאן >>