Uniformity in mechanical properties of the slab affects quality of subsequent rolling process. One of the most important factors deciding quality of the slab is fluctuation of the molten steel level in the mould. That is, smoothing pouring without fluctuating in the mould level means improvement in quality of the slab and protects break-out problem and allows high speed casting process. If molten steel surface fluctuates severely, the forming oscillation marks on the slab is unstable, solidification of molten steel is not uniform and there will be entrapment of mould powder in the solidified cast strand. It makes quality of the slab inferior and generates defects on the slab.
https://telecombcn-dl.github.io/idl-2020/
Deep learning technologies are at the core of the current revolution in artificial intelligence for multimedia data analysis. The convergence of large-scale annotated datasets and affordable GPU hardware has allowed the training of neural networks for data analysis tasks which were previously addressed with hand-crafted features. Architectures such as convolutional neural networks, recurrent neural networks or Q-nets for reinforcement learning have shaped a brand new scenario in signal processing. This course will cover the basic principles of deep learning from both an algorithmic and computational perspectives.
For the full video of this presentation, please visit: https://www.edge-ai-vision.com/2023/07/how-transformers-are-changing-the-nature-of-deep-learning-models-a-presentation-from-synopsys/
Tom Michiels, System Architect for ARC Processors at Synopsys, presents the “How Transformers Are Changing the Nature of Deep Learning Models” tutorial at the May 2023 Embedded Vision Summit.
The neural network models used in embedded real-time applications are evolving quickly. Transformer networks are a deep learning approach that has become dominant for natural language processing and other time-dependent, series data applications. Now, transformer-based deep learning network architectures are also being applied to vision applications with state-of-the-art results compared to CNN-based solutions.
In this presentation, Michiels introduces transformers and contrasts them with the CNNs commonly used for vision tasks today. He examines the key features of transformer model architectures and shows performance comparisons between transformers and CNNs. He concludes with insights on why his company thinks transformers will become increasingly important for visual perception tasks.
SLIDING WINDOW SUM ALGORITHMS FOR DEEP NEURAL NETWORKSIJCI JOURNAL
Sliding window sums are widely used for string indexing, hashing and time series analysis. We have
developed a family of the generic vectorized sliding sum algorithms that provide speedup of O(P/w) for
window size w and number of processors P. For a sum with a commutative operator the speedup is
improved to O(P/log(w)). Even more important, our algorithms exhibit efficient memory access patterns. In
this paper we study the application of sliding sum algorithms to the training and inference of Deep Neural
Networks. We demonstrate how both pooling and convolution primitives could be expressed as sliding
sums and evaluated by the compute kernels with a shared structure. We show that the sliding sum
convolution kernels are more efficient than the commonly used GEMM kernels on CPUs and could even
outperform their GPU counterparts.
Uniformity in mechanical properties of the slab affects quality of subsequent rolling process. One of the most important factors deciding quality of the slab is fluctuation of the molten steel level in the mould. That is, smoothing pouring without fluctuating in the mould level means improvement in quality of the slab and protects break-out problem and allows high speed casting process. If molten steel surface fluctuates severely, the forming oscillation marks on the slab is unstable, solidification of molten steel is not uniform and there will be entrapment of mould powder in the solidified cast strand. It makes quality of the slab inferior and generates defects on the slab.
https://telecombcn-dl.github.io/idl-2020/
Deep learning technologies are at the core of the current revolution in artificial intelligence for multimedia data analysis. The convergence of large-scale annotated datasets and affordable GPU hardware has allowed the training of neural networks for data analysis tasks which were previously addressed with hand-crafted features. Architectures such as convolutional neural networks, recurrent neural networks or Q-nets for reinforcement learning have shaped a brand new scenario in signal processing. This course will cover the basic principles of deep learning from both an algorithmic and computational perspectives.
For the full video of this presentation, please visit: https://www.edge-ai-vision.com/2023/07/how-transformers-are-changing-the-nature-of-deep-learning-models-a-presentation-from-synopsys/
Tom Michiels, System Architect for ARC Processors at Synopsys, presents the “How Transformers Are Changing the Nature of Deep Learning Models” tutorial at the May 2023 Embedded Vision Summit.
The neural network models used in embedded real-time applications are evolving quickly. Transformer networks are a deep learning approach that has become dominant for natural language processing and other time-dependent, series data applications. Now, transformer-based deep learning network architectures are also being applied to vision applications with state-of-the-art results compared to CNN-based solutions.
In this presentation, Michiels introduces transformers and contrasts them with the CNNs commonly used for vision tasks today. He examines the key features of transformer model architectures and shows performance comparisons between transformers and CNNs. He concludes with insights on why his company thinks transformers will become increasingly important for visual perception tasks.
SLIDING WINDOW SUM ALGORITHMS FOR DEEP NEURAL NETWORKSIJCI JOURNAL
Sliding window sums are widely used for string indexing, hashing and time series analysis. We have
developed a family of the generic vectorized sliding sum algorithms that provide speedup of O(P/w) for
window size w and number of processors P. For a sum with a commutative operator the speedup is
improved to O(P/log(w)). Even more important, our algorithms exhibit efficient memory access patterns. In
this paper we study the application of sliding sum algorithms to the training and inference of Deep Neural
Networks. We demonstrate how both pooling and convolution primitives could be expressed as sliding
sums and evaluated by the compute kernels with a shared structure. We show that the sliding sum
convolution kernels are more efficient than the commonly used GEMM kernels on CPUs and could even
outperform their GPU counterparts.
QUATERNARY LOGIC AND APPLICATIONS USING MULTIPLE QUANTUM WELL BASED SWSFETSVLSICS Design
This paper presents Spatial Wavefunction-Switched Field-Effect Transistors (SWSFET) to implement efficient quaternary logic and arithmetic functions. Various quaternary logic gates and digital building blocks are presented using SWSFETs. In addition, arithmetic operation with full adder using novel logic algebra is also presented. The SWSFET based implementation of digital logic, cache and arithmetic block results in up to 75% reduction in transistor count and up to 50% reduction in data interconnect densities. Simulations of quaternary logic gates using the BSIM equivalent models for SWSFET channels are also described.
TMPA-2017: Modeling of PLC-programs by High-level Coloured Petri NetsIosif Itkin
TMPA-2017: Tools and Methods of Program Analysis
3-4 March, 2017, Hotel Holiday Inn Moscow Vinogradovo, Moscow
Modeling of PLC-programs by High-level Coloured Petri Nets
Dmitriy Ryabukhin, Egor Kuzmin, Valery Sokolov, Yaroslavl State University
For video follow the link: https://youtu.be/XJoKuCNrTi0
Would like to know more?
Visit our website:
www.tmpaconf.org
www.exactprosystems.com/events/tmpa
Follow us:
https://www.linkedin.com/company/exactpro-systems-llc?trk=biz-companies-cym
https://twitter.com/exactpro
Computational steering Interactive Design-through-Analysis for Simulation Sci...SURFevents
Computational steering has evolved with advances in computing and visualization technologies. This session will showcase interactive design-through-analysis techniques that seamlessly integrate computer-aided design and simulation-based analysis tools. The approach replaces traditional simulation-based analysis with IgANets, which embeds physics-informed machine learning into the Isogeometric Analysis paradigm. IgANets train parametrized deep networks to predict solution coefficients of B-Spline/NURBS representations, enabling instantaneous evaluation and interactive feedback loops. A first-of-its-kind demonstrator coupling IgANets with a novel user frontend, developed at SURF, will be presented to initiate a new trend in computational steering towards interactive design-through-analysis.
Modelling Quantum Transport in Nanostructuresiosrjce
IOSR Journal of Electronics and Communication Engineering(IOSR-JECE) is a double blind peer reviewed International Journal that provides rapid publication (within a month) of articles in all areas of electronics and communication engineering and its applications. The journal welcomes publications of high quality papers on theoretical developments and practical applications in electronics and communication engineering. Original research papers, state-of-the-art reviews, and high quality technical notes are invited for publications.
Connectomics: Parcellations and Network Analysis MethodsGael Varoquaux
Simple tutorial on methods for functional connectome analysis: learning regions, extracting functional signal, inferring the network structure, and comparing it across subjects.
My presentation at University of Nottingham "Fast low-rank methods for solvin...Alexander Litvinenko
Overview of my (with co-authors) low-rank tensor methods for solving PDEs with uncertain coefficients. Connection with Bayesian Update. Solving a coupled system: stochastic forward and stochastic inverse.
Qubit models and methods for improving the performance of software and hardware for
analyzing digital devices through increasing the dimension of the data structures and memory
are proposed. The basic concepts, terminology and definitions necessary for the implementation
of quantum computing when analyzing virtual computers are introduced. The investigation results concerning design and modeling computer systems in a cyberspace based on the use of two-component structure <memory> are presented.
Richard's entangled aventures in wonderlandRichard Gill
Since the loophole-free Bell experiments of 2020 and the Nobel prizes in physics of 2022, critics of Bell's work have retreated to the fortress of super-determinism. Now, super-determinism is a derogatory word - it just means "determinism". Palmer, Hance and Hossenfelder argue that quantum mechanics and determinism are not incompatible, using a sophisticated mathematical construction based on a subtle thinning of allowed states and measurements in quantum mechanics, such that what is left appears to make Bell's argument fail, without altering the empirical predictions of quantum mechanics. I think however that it is a smoke screen, and the slogan "lost in math" comes to my mind. I will discuss some other recent disproofs of Bell's theorem using the language of causality based on causal graphs. Causal thinking is also central to law and justice. I will mention surprising connections to my work on serial killer nurse cases, in particular the Dutch case of Lucia de Berk and the current UK case of Lucy Letby.
QUATERNARY LOGIC AND APPLICATIONS USING MULTIPLE QUANTUM WELL BASED SWSFETSVLSICS Design
This paper presents Spatial Wavefunction-Switched Field-Effect Transistors (SWSFET) to implement efficient quaternary logic and arithmetic functions. Various quaternary logic gates and digital building blocks are presented using SWSFETs. In addition, arithmetic operation with full adder using novel logic algebra is also presented. The SWSFET based implementation of digital logic, cache and arithmetic block results in up to 75% reduction in transistor count and up to 50% reduction in data interconnect densities. Simulations of quaternary logic gates using the BSIM equivalent models for SWSFET channels are also described.
TMPA-2017: Modeling of PLC-programs by High-level Coloured Petri NetsIosif Itkin
TMPA-2017: Tools and Methods of Program Analysis
3-4 March, 2017, Hotel Holiday Inn Moscow Vinogradovo, Moscow
Modeling of PLC-programs by High-level Coloured Petri Nets
Dmitriy Ryabukhin, Egor Kuzmin, Valery Sokolov, Yaroslavl State University
For video follow the link: https://youtu.be/XJoKuCNrTi0
Would like to know more?
Visit our website:
www.tmpaconf.org
www.exactprosystems.com/events/tmpa
Follow us:
https://www.linkedin.com/company/exactpro-systems-llc?trk=biz-companies-cym
https://twitter.com/exactpro
Computational steering Interactive Design-through-Analysis for Simulation Sci...SURFevents
Computational steering has evolved with advances in computing and visualization technologies. This session will showcase interactive design-through-analysis techniques that seamlessly integrate computer-aided design and simulation-based analysis tools. The approach replaces traditional simulation-based analysis with IgANets, which embeds physics-informed machine learning into the Isogeometric Analysis paradigm. IgANets train parametrized deep networks to predict solution coefficients of B-Spline/NURBS representations, enabling instantaneous evaluation and interactive feedback loops. A first-of-its-kind demonstrator coupling IgANets with a novel user frontend, developed at SURF, will be presented to initiate a new trend in computational steering towards interactive design-through-analysis.
Modelling Quantum Transport in Nanostructuresiosrjce
IOSR Journal of Electronics and Communication Engineering(IOSR-JECE) is a double blind peer reviewed International Journal that provides rapid publication (within a month) of articles in all areas of electronics and communication engineering and its applications. The journal welcomes publications of high quality papers on theoretical developments and practical applications in electronics and communication engineering. Original research papers, state-of-the-art reviews, and high quality technical notes are invited for publications.
Connectomics: Parcellations and Network Analysis MethodsGael Varoquaux
Simple tutorial on methods for functional connectome analysis: learning regions, extracting functional signal, inferring the network structure, and comparing it across subjects.
My presentation at University of Nottingham "Fast low-rank methods for solvin...Alexander Litvinenko
Overview of my (with co-authors) low-rank tensor methods for solving PDEs with uncertain coefficients. Connection with Bayesian Update. Solving a coupled system: stochastic forward and stochastic inverse.
Qubit models and methods for improving the performance of software and hardware for
analyzing digital devices through increasing the dimension of the data structures and memory
are proposed. The basic concepts, terminology and definitions necessary for the implementation
of quantum computing when analyzing virtual computers are introduced. The investigation results concerning design and modeling computer systems in a cyberspace based on the use of two-component structure <memory> are presented.
Richard's entangled aventures in wonderlandRichard Gill
Since the loophole-free Bell experiments of 2020 and the Nobel prizes in physics of 2022, critics of Bell's work have retreated to the fortress of super-determinism. Now, super-determinism is a derogatory word - it just means "determinism". Palmer, Hance and Hossenfelder argue that quantum mechanics and determinism are not incompatible, using a sophisticated mathematical construction based on a subtle thinning of allowed states and measurements in quantum mechanics, such that what is left appears to make Bell's argument fail, without altering the empirical predictions of quantum mechanics. I think however that it is a smoke screen, and the slogan "lost in math" comes to my mind. I will discuss some other recent disproofs of Bell's theorem using the language of causality based on causal graphs. Causal thinking is also central to law and justice. I will mention surprising connections to my work on serial killer nurse cases, in particular the Dutch case of Lucia de Berk and the current UK case of Lucy Letby.
Observation of Io’s Resurfacing via Plume Deposition Using Ground-based Adapt...Sérgio Sacani
Since volcanic activity was first discovered on Io from Voyager images in 1979, changes
on Io’s surface have been monitored from both spacecraft and ground-based telescopes.
Here, we present the highest spatial resolution images of Io ever obtained from a groundbased telescope. These images, acquired by the SHARK-VIS instrument on the Large
Binocular Telescope, show evidence of a major resurfacing event on Io’s trailing hemisphere. When compared to the most recent spacecraft images, the SHARK-VIS images
show that a plume deposit from a powerful eruption at Pillan Patera has covered part
of the long-lived Pele plume deposit. Although this type of resurfacing event may be common on Io, few have been detected due to the rarity of spacecraft visits and the previously low spatial resolution available from Earth-based telescopes. The SHARK-VIS instrument ushers in a new era of high resolution imaging of Io’s surface using adaptive
optics at visible wavelengths.
Earliest Galaxies in the JADES Origins Field: Luminosity Function and Cosmic ...Sérgio Sacani
We characterize the earliest galaxy population in the JADES Origins Field (JOF), the deepest
imaging field observed with JWST. We make use of the ancillary Hubble optical images (5 filters
spanning 0.4−0.9µm) and novel JWST images with 14 filters spanning 0.8−5µm, including 7 mediumband filters, and reaching total exposure times of up to 46 hours per filter. We combine all our data
at > 2.3µm to construct an ultradeep image, reaching as deep as ≈ 31.4 AB mag in the stack and
30.3-31.0 AB mag (5σ, r = 0.1” circular aperture) in individual filters. We measure photometric
redshifts and use robust selection criteria to identify a sample of eight galaxy candidates at redshifts
z = 11.5 − 15. These objects show compact half-light radii of R1/2 ∼ 50 − 200pc, stellar masses of
M⋆ ∼ 107−108M⊙, and star-formation rates of SFR ∼ 0.1−1 M⊙ yr−1
. Our search finds no candidates
at 15 < z < 20, placing upper limits at these redshifts. We develop a forward modeling approach to
infer the properties of the evolving luminosity function without binning in redshift or luminosity that
marginalizes over the photometric redshift uncertainty of our candidate galaxies and incorporates the
impact of non-detections. We find a z = 12 luminosity function in good agreement with prior results,
and that the luminosity function normalization and UV luminosity density decline by a factor of ∼ 2.5
from z = 12 to z = 14. We discuss the possible implications of our results in the context of theoretical
models for evolution of the dark matter halo mass function.
Cancer cell metabolism: special Reference to Lactate PathwayAADYARAJPANDEY1
Normal Cell Metabolism:
Cellular respiration describes the series of steps that cells use to break down sugar and other chemicals to get the energy we need to function.
Energy is stored in the bonds of glucose and when glucose is broken down, much of that energy is released.
Cell utilize energy in the form of ATP.
The first step of respiration is called glycolysis. In a series of steps, glycolysis breaks glucose into two smaller molecules - a chemical called pyruvate. A small amount of ATP is formed during this process.
Most healthy cells continue the breakdown in a second process, called the Kreb's cycle. The Kreb's cycle allows cells to “burn” the pyruvates made in glycolysis to get more ATP.
The last step in the breakdown of glucose is called oxidative phosphorylation (Ox-Phos).
It takes place in specialized cell structures called mitochondria. This process produces a large amount of ATP. Importantly, cells need oxygen to complete oxidative phosphorylation.
If a cell completes only glycolysis, only 2 molecules of ATP are made per glucose. However, if the cell completes the entire respiration process (glycolysis - Kreb's - oxidative phosphorylation), about 36 molecules of ATP are created, giving it much more energy to use.
IN CANCER CELL:
Unlike healthy cells that "burn" the entire molecule of sugar to capture a large amount of energy as ATP, cancer cells are wasteful.
Cancer cells only partially break down sugar molecules. They overuse the first step of respiration, glycolysis. They frequently do not complete the second step, oxidative phosphorylation.
This results in only 2 molecules of ATP per each glucose molecule instead of the 36 or so ATPs healthy cells gain. As a result, cancer cells need to use a lot more sugar molecules to get enough energy to survive.
Unlike healthy cells that "burn" the entire molecule of sugar to capture a large amount of energy as ATP, cancer cells are wasteful.
Cancer cells only partially break down sugar molecules. They overuse the first step of respiration, glycolysis. They frequently do not complete the second step, oxidative phosphorylation.
This results in only 2 molecules of ATP per each glucose molecule instead of the 36 or so ATPs healthy cells gain. As a result, cancer cells need to use a lot more sugar molecules to get enough energy to survive.
introduction to WARBERG PHENOMENA:
WARBURG EFFECT Usually, cancer cells are highly glycolytic (glucose addiction) and take up more glucose than do normal cells from outside.
Otto Heinrich Warburg (; 8 October 1883 – 1 August 1970) In 1931 was awarded the Nobel Prize in Physiology for his "discovery of the nature and mode of action of the respiratory enzyme.
WARNBURG EFFECT : cancer cells under aerobic (well-oxygenated) conditions to metabolize glucose to lactate (aerobic glycolysis) is known as the Warburg effect. Warburg made the observation that tumor slices consume glucose and secrete lactate at a higher rate than normal tissues.
Multi-source connectivity as the driver of solar wind variability in the heli...Sérgio Sacani
The ambient solar wind that flls the heliosphere originates from multiple
sources in the solar corona and is highly structured. It is often described
as high-speed, relatively homogeneous, plasma streams from coronal
holes and slow-speed, highly variable, streams whose source regions are
under debate. A key goal of ESA/NASA’s Solar Orbiter mission is to identify
solar wind sources and understand what drives the complexity seen in the
heliosphere. By combining magnetic feld modelling and spectroscopic
techniques with high-resolution observations and measurements, we show
that the solar wind variability detected in situ by Solar Orbiter in March
2022 is driven by spatio-temporal changes in the magnetic connectivity to
multiple sources in the solar atmosphere. The magnetic feld footpoints
connected to the spacecraft moved from the boundaries of a coronal hole
to one active region (12961) and then across to another region (12957). This
is refected in the in situ measurements, which show the transition from fast
to highly Alfvénic then to slow solar wind that is disrupted by the arrival of
a coronal mass ejection. Our results describe solar wind variability at 0.5 au
but are applicable to near-Earth observatories.
Seminar of U.V. Spectroscopy by SAMIR PANDASAMIR PANDA
Spectroscopy is a branch of science dealing the study of interaction of electromagnetic radiation with matter.
Ultraviolet-visible spectroscopy refers to absorption spectroscopy or reflect spectroscopy in the UV-VIS spectral region.
Ultraviolet-visible spectroscopy is an analytical method that can measure the amount of light received by the analyte.
Unveiling the Energy Potential of Marshmallow Deposits.pdf
Heuristic Approach for Model Reduction of Large-Scale Logistics Networks
1. Heuristic Approach for Model Reduction of
Large-Scale Logistics Networks
Michael Kosmykov
Centre for Industrial Mathematics, University of Bremen, Germany
February 14, 2011, Elgersburg Workshop 2011, Elgersburg
joint work with Sergey Dashkovskiy, Thomas Makuschewitz,
Bernd Scholz-Reiter, Michael Sch¨onlein and Fabian Wirth
2. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Outline
1 Motivation
2 Network structure information
3 Structure preserving model reduction
4 Application (Jackson networks)
5 Conclusions and outlook
2 / 30Motivation Structure Reduction Application Conclusions
3. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Outline
1 Motivation
2 Network structure information
3 Structure preserving model reduction
4 Application (Jackson networks)
5 Conclusions and outlook
3 / 30Motivation Structure Reduction Application Conclusions
4. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Characteristics of the practical example:
3 production sites (D, F, E) for
Liquidring-Vaccum (LRVP), Industrial (IND)
and Side-channel (SC) pumps
5 distribution centers (D, NL, B, F, E)
33 first and second-tier suppliers for the
production of pumps
90 suppliers for components that are needed
for the assembly of pump sets
More than 1000 customers
4 / 30Motivation Structure Reduction Application Conclusions
5. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Analysis steps:
1 Mathematical modelling
2 Model reduction
3 Analysis of the reduced model
4 Analysis of relationships between reduced and original models
Main methods of model reduction:
Balancing methods
Krylov methods
Drawback of application to logistics networks: the interconnection
structure of the network is destroyed
5 / 30Motivation Structure Reduction Application Conclusions
6. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Structure-preserving model reduction
Advantages for logistics networks:
keeps the physical meaning of the objects
allows analysis of interrelationships between locations
allows identification of influential logistic objects
Our heuristic approach uses
importance of locations for the network
structure of material flows in the network
6 / 30Motivation Structure Reduction Application Conclusions
7. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Outline
1 Motivation
2 Network structure information
3 Structure preserving model reduction
4 Application (Jackson networks)
5 Conclusions and outlook
7 / 30Motivation Structure Reduction Application Conclusions
8. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Graph of logistics network
f11,16
v1v1v1 v2v2v2 v3v3v3 v4v4v4 v5v5v5 v6v6v6
v7v7v7 v8v8v8 v9v9v9 v10v10v10
v11v11v11 v12v12v12 v13v13v13 v14v14v14 v15v15v15
v16v16v16 v17v17v17
v18v18v18 v19v19v19
v20v20v20 v21v21v21 v22v22v22 v23v23v23
fij - material flow from location i to j
8 / 30Motivation Structure Reduction Application Conclusions
10. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Ranking
mij - probability to move from vertex j to vertex i
xi (k) - probability that at kth step an order is placed at location i
Markov chain: x(k + 1) = Mx(k)
Rank ri (importance) of location i := probability that a random
order is placed location i = stationary distribution of Markov chain:
Mr = r, r ∈ Rn
.
Perron-Frobenius Theorem ⇒ if the transition matrix is irreducible
the Markov chain has a unique stationary distribution r > 0.
10 / 30Motivation Structure Reduction Application Conclusions
11. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Graph modification
Connecting retailers and source suppliers:
v1v1v1 v2v2v2
v3v3v3 v4v4v4
v5v5v5 v6v6v6 v7v7v7
15
20 5
50
10
50
Material flow between the locations (E)
Relative capacity allocation (E )
New edge set = E∪E
S - source suppliers
Rj - retailers affecting supplier j
P(i) - locations connected to retailer i
pi :=
k∈P(i)
fki , i ∈ Rj
qj :=
i∈Rj
pi , j ∈ S
Mij :=
mij (i, j) ∈ E,
pi
qj
(i, j) ∈ E ,
0 else.
11 / 30Motivation Structure Reduction Application Conclusions
13. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Embedding into a larger network
c ∈ [0, 1] - the strength of connection
v = [vT
n vT
m]T
, w = [wT
n wT
m]T
, n+m n
- connection with an outside world
L =
cM + (1 − c)wneT
n vneT
m
(1 − c)wmeT
n vmeT
m
12 / 30Motivation Structure Reduction Application Conclusions
14. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
LogRank
Theorem
For v > 0, wm > 0 and the Perron vector r partitioned as [rT
n rT
m]T
rn is an eigenvector corresponding to the eigenvalue 1 of
Mc(v, w) := cM + (1 − c) wn +
eT
mwm
1 − eT
mvm
vn eT
n .
Furthermore, Mc(v, w) is primitive.
LogRank
The normalized eigenvector associated to the eigenvalue 1 of
Mc(v, w) is called the LogRank.
Idea is similar to the PageRank by Page et al. 1998.
13 / 30Motivation Structure Reduction Application Conclusions
18. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Aggregation of vertices
Theorem
Consider a strongly connected weighted directed graph
G = (V , E, M). Let r be the (unique) normalized Perron vector of
M. Given a disjoint partition
V = {1, . . . , n} =: J1 ∪ J2 ∪ . . . ∪ Jk , with ˜V = {J1, . . . , Jk},
˜mij := ν∈Ji
1
µ∈Jj
rµ µ∈Jj
rµmνµ , then ˜M is irreducible and the
unique normalized Perron vector ˜r of ˜M has the property
˜ri =
ν∈Ji
rν , i = 1, . . . , k . (1)
17 / 30Motivation Structure Reduction Application Conclusions
19. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Aggregation of certain motifs
Parallel connections:
vvv
v1v1v1 vkvkvk
vvv
... ⇒
vvv
JJJ
vvv
The LogRank of vJ ∈ ˜V is the sum of the LogRanks of the aggregated
vertices in VJ , while the LogRank of the unaffected vertices v1, . . . , vl+2
is preserved. That is,
˜rT
= r1 . . . rl+2 rn−k+1 + . . . + rn .
18 / 30Motivation Structure Reduction Application Conclusions
20. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Aggregation of certain motifs
Sequential connections:
vvv v1v1v1 ......... vkvkvk vvv ⇒ vvv JJJ vvv
The LogRank of vJ ∈ ˜V is the sum of the LogRanks of the aggregated
vertices in VJ , while the LogRank of the unaffected vertices v1, . . . , vl+2
is preserved. That is,
˜rT
= r1 . . . rl+2 rn−k+1 + . . . + rn .
19 / 30Motivation Structure Reduction Application Conclusions
21. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Aggregation of certain motifs
Almost disconnected subgraphs:
v1v1v1
v2v2v2
v3v3v3
v∗v∗
v∗
⇒
JJJ
The LogRank of vJ ∈ ˜V is the sum of the LogRanks of the aggregated
vertices in VJ , while the LogRank of the unaffected vertices v1, . . . , vl+2
is preserved. That is,
˜rT
= r1 . . . rl+2 rn−k+1 + . . . + rn .
20 / 30Motivation Structure Reduction Application Conclusions
22. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Analysis steps:
1 Modelling
2 Model reduction
Step 1. Calculation of ranks
Step 2. Identification of motifs with low rank locations
Step 3. Aggregation of motifs
3 Analysis of the reduced model
21 / 30Motivation Structure Reduction Application Conclusions
23. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Meta algorithm :
Compute the LogRank r of the network G = (V , E, Mc (v, w)) and
generate R∆
repeat Delete and consider v1 ∈ R∆;
Generate candidate list C = C(v1, r);
whileC = ∅
Delete and consider c1 from the candidate list C;
if for c1 reduction error e ≤ ε
aggregate c1;
clear C;
Generate new waiting list R∆;
end if
end while
untilR∆ = ∅
22 / 30Motivation Structure Reduction Application Conclusions
24. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Outline
1 Motivation
2 Network structure information
3 Structure preserving model reduction
4 Application (Jackson networks)
5 Conclusions and outlook
23 / 30Motivation Structure Reduction Application Conclusions
25. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Example: Jackson networks
Open Jackson network is a queueing network where:
procesing time at each location is i.i.d. with exponential
distribution
all orders belong to the same class, follow the same service
time distribution and the same routing mechanism
orders arrive from the outside according to the Poisson
process with rate α > 0
each arriving order is independently routed to location i with
probability p0i ≥ 0
after being processed at some location an order routes
according to routing matrix P
24 / 30Motivation Structure Reduction Application Conclusions
26. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Dynamics of Jackson networks
The effective arrival rate λi of orders at location i is given by
traffic equation:
λi = α p0i +
n
j=1
pji λj , i ∈ V .
P :=
PT pe
pT
o 0
.
pe - external inflow probability vector
po - outflow probability vector
If P is irreducible, the LogRank and the effective arrival rate
coincide (up to a constant multiple).
25 / 30Motivation Structure Reduction Application Conclusions
27. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Example of reduction
v1v1v1 v2v2v2 v3v3v3 v4v4v4 v5v5v5 v6v6v6
v7v7v7 v8v8v8 v9v9v9 v10v10v10
v11v11v11 v12v12v12 v13v13v13 v14v14v14 v15v15v15
v16v16v16 v17v17v17
v18v18v18 v19v19v19
v20v20v20 v21v21v21 v22v22v22 v23v23v23
High rank
Low rank
26 / 30Motivation Structure Reduction Application Conclusions
28. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Example of reduction
v2v2v2 v3v3v3 v5v5v5
v7v7v7
v8v8v8 v9v9v9
v11v11v11 v12v12v12 v14v14v14
v16v16v16 v17v17v17
v18v18v18 v19v19v19
v20v20v20 v21v21v21 v23v23v23
High rank
Low rank
27 / 30Motivation Structure Reduction Application Conclusions
29. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Outline
1 Motivation
2 Network structure information
3 Structure preserving model reduction
4 Application (Jackson networks)
5 Conclusions and outlook
28 / 30Motivation Structure Reduction Application Conclusions
30. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Conclusions
A new heuristic approach for structure-preserving model
reduction based on:
material flow information
certain motifs of the network
rank of importance of locations
Application to Jackson networks
Further research:
Preservation of stability properties
Estimation of approximation error
Application to networks with nonlinear dynamics
29 / 30Motivation Structure Reduction Application Conclusions
31. Centre for
Industrial Mathematics
Elgersburg Workshop 2011
Kosmykov
Thank you for your attention!
This work is part of the research project: ”Stability, Robustness and
Approximation of Dynamic Large-Scale Networks - Theory and
Applications in Logistics Networks” funded by the Volkswagen
Foundation.
30 / 30Motivation Structure Reduction Application Conclusions