Nonnegative/Binary matrix factorization for image classification … – Nature.com
NBMF9 extracts features by decomposing data into basis and coefficient matrices. The dataset was converted into a positive matrix to prepare the input for the NBMF method. If the dataset contains n-dimensional m data, then the input is an (n times m) matrix V. The input matrix is decomposed into a base matrix W of (n times k), representing the dataset features, and a coefficient matrix H of (k times m), representing the combination of features selected to reconstruct the original matrix. Then,
$$begin{aligned} V approx WH, end{aligned}$$
(1)
where W and H are positive and binary matrices, respectively. The column number k of W corresponds to the number of features extracted from the data and can be set to any value. To minimize the difference between the left and right sides of Eq.(1), W and H are updated alternately as
$$ W: = mathop {{text{arg}}~{text{min}}}limits_{{X in mathbb{R}^{{ + n times k}} }} parallel V - XHparallel _{F} + alpha parallel Xparallel _{F} , $$
(2)
$$ H: = mathop {{text{arg}}~{text{min}}}limits_{{X in { 0,1} ^{{k times m}} }} parallel V - WXparallel _{F} , $$
(3)
where (parallel cdot parallel _F) denotes the Frobenius norm. The components of W and H are initially given randomly. The hyperparameter (alpha ) is a positive real value that prevents overfitting and is set to (alpha = 1.0 times 10^{-4}).
In previous studies, the Projected Gradient Method (PGM) was used to update Eq.(2)16. The loss function that updates Eq.(2) is defined as
$$begin{aligned} f_{W}(varvec{x}) = parallel varvec{v} - H^{text{T}} varvec{x} parallel ^{2} + alpha parallel varvec{x} parallel ^{2}, end{aligned}$$
(4)
where (varvec{x}^{text{T}}) and (varvec{v}^{text{T}}) are the row vectors of W and V, respectively. The gradient of Eq.(4) is expressed as
$$begin{aligned} nabla f_{W} = -H (varvec{v} - H^T varvec{x}) + alpha varvec{x}. end{aligned}$$
(5)
The PGM minimizes the loss functions in Eq.(4) by updating (varvec{x}):
$$begin{aligned} varvec{x}^{t+1} = Pleft[varvec{x}^t - gamma _t nabla f_W (varvec{x}^t)right], end{aligned}$$
(6)
where (gamma _t) is the learning rate and
$$begin{aligned} P[x_i] = {left{ begin{array}{ll} 0 &{} (x_i le 0), \ x_i &{} (0< x_i < x_mathrm{{max}}), \ x_mathrm{{max}} &{} (x_mathrm{{max}} le x_i), end{array}right. } end{aligned}$$
(7)
where (x_mathrm{{max}}) is the upper bound and is set to (x_mathrm{{max}}=1). Eq.(7) is a projection that keeps the components of (varvec{x}) nonnegative.
However, because H is a binary matrix, Eq.(3) can be regarded as a combinatorial optimization problem that can be minimized by using an annealing method. To solve Eq.(3) using a D-Wave machine, a quantum annealing computer, we formulated the loss function as a quadratic unconstrained binary optimization model:
$$begin{aligned} f_{H}(varvec{q}) = sum _i sum _r W_{ri}left(W_{ri} - 2 v_{r}right) q_i + 2 sum _{i
(8)
where (varvec{q}) and (varvec{v}) are the column vectors of H and V, respectively.
After the alternate updating method converges, we obtain W and H which minimize the difference between the left and right sides of Eq.(1). W consists of representative features extracted from the input data, and H represents the combination of features in W using binary values to reconstruct V. Therefore, V can be approximated as the product of W and H.
Previous studies used NBMF to extract features from facial images9. When the number of annealing steps is small, the computation time is shorter than a classical combinatorial optimization solver. However, using the D-Wave machine is disadvantageous in that the computing time increases linearly with the number of annealings, whereas the classical solver does not significantly change the computing time. The results were compared with NMF14. Unlike NBMF, matrix H in NMF is positive and not binary. While the matrix H produced by NBMF was sparser than NMF, the difference between V and WH of NBMF was approximately 2.17 times larger than NMF. Although NBMF can have a shorter data processing time than the classical method, it is inferior to NMF as a machine-learning method in accuracy. Moreover, because previous studies did not demonstrate tasks beyond data reconstruction, the usefulness of NBMF as a machine-learning model is uncertain.
In this study, we propose the application of NBMF to a multiclass classification model. Inspired by the structure of a fully connected neural network (FCNN), we define an image classification model using NBMF. In an FCNN, image data are fed into the network as input, as shown in Fig.1, and the predicted classes are obtained as the output of the network through the hidden layers.
An overview of a fully-connected neural network.
To perform fully connected network learning using NBMF, we interpret the structure shown in Fig.1 as a single-matrix decomposition. When the input and output layers of the FCNN are combined into one input layer, the network becomes a two-layer network with the same structure as NBMF. As the input to the training network by NBMF, we used a matrix consisting of image data and the corresponding class information. Class information is represented by a one-hot vector multiplied by an arbitrary real number g. The image data and class information vectors are combined row-wise and eventually transformed into an input matrix V. We use NBMF to decompose V to obtain the basis matrix W and the coefficient matrix H, as shown in Fig.2. The column vectors in H correspond to the nodes in the hidden layer of the FCNN network, and the components of W correspond to the weights of the edges. The number of feature dimensions k in the NBMF corresponds to the number of nodes in the hidden layer of the FCNN.
An overview of training by NBMF.
To obtain H, we minimize Eq.(8) by using an annealing solver, as in a previous study. However, to obtain W by minimizing Eq.(4), we propose using the projected Root Mean Square Propagation (RMSProp) method instead of the PGM used in a previous study. RMSProp is a gradient descent method that adjusts the learning and decay rates to help the solution escape local minima17. RMSProp updates the vector (varvec{h}), whose components are denoted by (h_i) as
$$begin{aligned} h^{t+1}_{i} = beta h^{t}_{i} + (1-beta ) g^{2}_{i}, end{aligned}$$
(9)
where (beta ) is the decay rate, (varvec{g} = nabla f_W), and vector (varvec{x}) is
$$begin{aligned} varvec{x}^{t+1} = varvec{x}^{t} - eta frac{1}{sqrt{varvec{h}^{t} + epsilon }} nabla f_{W}, end{aligned}$$
(10)
where (eta ) is the learning rate, and (epsilon ) is a small value that prevents computational errors. After updating (varvec{x}) using Eq.(10), we apply the projections described in Eq.(7), to ensure that the solution does not exceed the bounds. We propose this method as a projected RMSProp.
In Fig.3, we demonstrated the information contained in W. Because the row vectors of W correspond to those of V, W consists of (W_1) corresponding to the image data information, and (W_2) corresponding to the class information. We plotted four column vectors selected from W trained with MNIST handwritten digit images under the conditions (m = 300) and (k=40), as shown in Fig.3. The images in Fig.3 show the column vectors of (W_1). The blue histograms show the frequencies at which the column vectors were selected to reconstruct the training data images with each label. The orange bar graphs show the component values of the corresponding column vectors of (W_2). For example, the image in Fig.3a resembles Number 0. From the histogram next to the image, we understand that the image is often used to reconstruct the training data labeled as 0. In the bar graph on the right, the corresponding column vector of (W_2) has the largest component value at an index of 0. This indicates that the column vector corresponding to the image has a feature of Number 0. Similarly, the image in Fig.3b has a label of 9. However, the image in Fig.3c appears to have curved features. From the histogram and bar graph next to the image, it appears that the image is often used to represent labels 2 and 3. This result is consistent with the fact that both numbers have a curve, which explains why the column vector of (W_1) was used in the reconstruction of images with labels 2 and 3. The image in Fig.3d has the shape of a straight line, and the corresponding histogram shows that the image is mainly used to express label 1 and is also frequently used to express label 6. Because Number 6 has a straight-line part, the result is reasonable.
The figure shows four sets of images, (a), (b), (c), and (d), corresponding to column vectors selected from W. Each set contains an image, a histogram, and a bar graph. The image represents a column vector of (W_1), and the histogram shows how often the column vector was selected to reconstruct the training data images with each label. The orange bar graph plots the component values of the corresponding column vector of (W_2).
In our multiclass classification model using NBMF, we used the trained matrices (W_1) and (W_2) to classify the test data in the workflow shown in Fig. 4.
An overview of testing by NBMF.
First, we decompose the test data matrix (V_text{test}) to obtain (H_text{test}) by using (W_1). Here, M represents the amount of test data, which corresponds to the number of column vectors of (V_text{test}). We use Eq.(3) for decomposition. Each column vector of (H_text{test}) represents the features selected from the trained (W_1) to approximate the corresponding column vector of (V_text{test}). Second, we multiply (W_2) by (H_text{test}) to obtain (U_text{test}), which expresses the prediction of the class vector corresponding to each column vector in (V_text{test}). Finally, we applied the softmax function to the components of (U_text{test}) and considered the index with the largest component value in each column vector to be the predicted class.
View original post here:
Nonnegative/Binary matrix factorization for image classification ... - Nature.com
- What's Going On With IonQ Stock Today? - Benzinga - April 1st, 2025 [April 1st, 2025]
- Quantum computer solves optimization problem at Ford's assembly line - Interesting Engineering - April 1st, 2025 [April 1st, 2025]
- Finnish Quantum Startup IQM in Talks to Raise Over 200 Million - Bloomberg.com - April 1st, 2025 [April 1st, 2025]
- Quantum Computing Approach Generates First Ever Truly Random Number - Discover Magazine - April 1st, 2025 [April 1st, 2025]
- National Quantum Computing Centre Launches Insights Paper Exploring Quantum Computings Transformative Potential in Healthcare and Pharmaceuticals -... - April 1st, 2025 [April 1st, 2025]
- JPMorganChase, Quantinuum, Argonne National Laboratory, Oak Ridge National Laboratory and University of Texas at Austin advance the application of... - April 1st, 2025 [April 1st, 2025]
- Certified randomness using a trapped-ion quantum processor - Nature - April 1st, 2025 [April 1st, 2025]
- What's Going On With Quantum Computing Stock Today? - Benzinga - April 1st, 2025 [April 1st, 2025]
- D-Wave Pushes Back At Critics, Shows Off Aggressive Quantum Roadmap - The Next Platform - April 1st, 2025 [April 1st, 2025]
- Quantum Computing Inc. Secures Quantum Photonic Vibrometer Order with Delft University of Technology - Yahoo Finance - April 1st, 2025 [April 1st, 2025]
- How quantum cybersecurity changes the way you protect data - TechTarget - April 1st, 2025 [April 1st, 2025]
- Pasqal Selected for 140-Qubit Quantum Computer to Be Hosted at CINECA - insideHPC - April 1st, 2025 [April 1st, 2025]
- D-Wave and Japan Tobacco use quantum to build a better AI model for drug discovery - SiliconANGLE - April 1st, 2025 [April 1st, 2025]
- Quantum Computing is a cross industry revolution, and we want to be part of it - CTech - April 1st, 2025 [April 1st, 2025]
- Quantum Computing Stocks Fall. Here's A Look At Upcoming News Events. - Investor's Business Daily - April 1st, 2025 [April 1st, 2025]
- Honeywell May Take Quantinuum Public in Next 2 Years. Its a Quantum Thing. - Barron's - April 1st, 2025 [April 1st, 2025]
- The 6 different types of quantum computing technology - TechTarget - April 1st, 2025 [April 1st, 2025]
- Nvidia to Open Quantum Computing Research Center in Boston This Year in a Landmark for Regions Tech Sector - The Harvard Crimson - April 1st, 2025 [April 1st, 2025]
- Quantum Threats Are HereWhy the Next Cybersecurity Boom May Already Be Underway - Baystreet.ca - April 1st, 2025 [April 1st, 2025]
- D-Wave and Japan Tobacco Validate Quantum and AI Workflow Towards Generative Drug Discovery - The Quantum Insider - April 1st, 2025 [April 1st, 2025]
- The High Cost of Quantum Randomness Is Dropping - Quanta Magazine - April 1st, 2025 [April 1st, 2025]
- Beyond encryption: Why quantum computing might be more of a science boom than a cybersecurity bust - oodaloop.com - April 1st, 2025 [April 1st, 2025]
- NVIDIA (NVDA): One of the Best Quantum Computing Stocks to Buy Right Now? - Yahoo Finance - March 18th, 2025 [March 18th, 2025]
- I work at a leading quantum lab: Here are the qualifications recruiters in the field are looking for - Business Insider - March 18th, 2025 [March 18th, 2025]
- 5 wild things quantum computing could unlock now that Big Tech believes a breakthrough is within reach - Yahoo - March 18th, 2025 [March 18th, 2025]
- Controversy erupts over claims Microsoft invented a new state of matter - Salon - March 18th, 2025 [March 18th, 2025]
- Chinese quantum processor is 1 quadrillion times faster than the best supercomputer and it rivals Google's breakthrough Willow chip - Livescience.com - March 18th, 2025 [March 18th, 2025]
- IQM Quantum wants to be the European answer to Google and IBM - Sifted - March 18th, 2025 [March 18th, 2025]
- Twisting atomically thin materials could advance quantum computers - University of Rochester - March 18th, 2025 [March 18th, 2025]
- D-Wave Quantum Stock Hits $11: Heres What This Top Analyst Predicts Ahead - TipRanks - March 18th, 2025 [March 18th, 2025]
- A Computer Has Achieved "Quantum Supremacy" On Real-World Problem For First Time, Company Claims - IFLScience - March 18th, 2025 [March 18th, 2025]
- INVESTOR ALERT: Pomerantz Law Firm Announces the Filing of a Class Action Against Quantum Computing Inc. and Certain Officers - QUBT - PR Newswire - March 18th, 2025 [March 18th, 2025]
- D-Wave Quantum Sets Benchmark with New Computing Advance - News and Statistics - IndexBox, Inc. - March 18th, 2025 [March 18th, 2025]
- Rigettis Rally Hits a Bump Are Insider Sales a Red Flag? - Wall Street Pit - March 18th, 2025 [March 18th, 2025]
- Quantum AI: What Is It and How Does It Work? - CNET - March 18th, 2025 [March 18th, 2025]
- D-Wave Shares Jump 46.9% on Friday - Should You Buy QBTS Stock? - TradingView - March 18th, 2025 [March 18th, 2025]
- 2 Top Quantum Computing Stocks to Buy in 2025 - The Motley Fool - March 13th, 2025 [March 13th, 2025]
- D-Wave Claims Breakthrough. Quantum Computing Stocks Gain. - Investor's Business Daily - March 13th, 2025 [March 13th, 2025]
- Physicists Just Witnessed a Quantum Phase Flip and Its More Mind-Bending Than Expected - SciTechDaily - March 13th, 2025 [March 13th, 2025]
- Beyond Classical: D-Wave First to Demonstrate Quantum Supremacy on Useful, Real-World Problem - Business Wire - March 13th, 2025 [March 13th, 2025]
- What is quantum computing and how it could change the tech world - Yahoo Finance - March 13th, 2025 [March 13th, 2025]
- Quantum Computing Giant IonQ Is Down More Than 60% From its All-Time High. Should You Buy The Dip? - The Motley Fool - March 13th, 2025 [March 13th, 2025]
- D-Wave Deep Dive: A Look at The Quantum Advantage Findings -- And The Questions That Remain - The Quantum Insider - March 13th, 2025 [March 13th, 2025]
- D-Wave claims to have achieved quantum supremacy at last, but others disagree - SiliconANGLE News - March 13th, 2025 [March 13th, 2025]
- D-Wave Claims It Achieves Quantum Supremacy. What the Breakthrough Means for Quantum Computing. - Barron's - March 13th, 2025 [March 13th, 2025]
- D-Wave Posts Wider-Than-Expected Loss. Why the Stock Is Rising After Earnings. - Barron's - March 13th, 2025 [March 13th, 2025]
- Nu Quantum Partners With The University of Sussex, Cisco, and Infineon to Scale Trapped Ion Quantum Computers - The Quantum Insider - March 13th, 2025 [March 13th, 2025]
- IonQ Could Be a Quantum Computing Powerhouse, but Is It a Buy Right Now? - The Motley Fool - March 13th, 2025 [March 13th, 2025]
- D-Wave Quantum Inc. Reports Fourth Quarter and Year-End 2024 Results - TradingView - March 13th, 2025 [March 13th, 2025]
- Hybrid Quantum Workflow Moves Toward Real-World Applications - IoT World Today - March 13th, 2025 [March 13th, 2025]
- As NVIDIAs Quantum Day Nears, Analysts Suggest Event is More Than a Gesture - The Quantum Insider - March 13th, 2025 [March 13th, 2025]
- D-Wave Posts Wider-Than-Expected Loss. Why the Stock Is Rising Anyway. - MSN - March 13th, 2025 [March 13th, 2025]
- Recent Breakthroughs Accelerate The Race For Quantum Computing - Forbes - March 13th, 2025 [March 13th, 2025]
- An operating system for executing applications on quantum network nodes - Nature.com - March 13th, 2025 [March 13th, 2025]
- D-Wave Reports Quantum Advantage in Materials Simulation Study - HPCwire - March 13th, 2025 [March 13th, 2025]
- Experts Weigh in on Microsofts Topological Qubit Claim - Physics - March 13th, 2025 [March 13th, 2025]
- Quantum Computing Inc. To Attend 37th Annual ROTH Conference - PR Newswire - March 13th, 2025 [March 13th, 2025]
- Quantum leap: Passwords in the new era of computing security - BleepingComputer - March 13th, 2025 [March 13th, 2025]
- Quantum computing will reach its inflection point in 2029: How investors should prepare - Finextra - March 13th, 2025 [March 13th, 2025]
- Quantum computing - Unlocking science, and maybe your bank account - Home Team Science and Technology Agency - March 13th, 2025 [March 13th, 2025]
- SXSW 2025 live coverage: The potential of quantum computing, Ireland's prime minister makes a splash, and a Metallica concert in Apple Vision Pro -... - March 13th, 2025 [March 13th, 2025]
- QuamCore Emerges From Stealth With $9 Million in Seed Funding to Build Worlds First Scalable 1 million Qubit Quantum Computer - The Quantum Insider - March 13th, 2025 [March 13th, 2025]
- QuamCore Emerges with $9M Seed Funding to Build Scalable Million-Qubit Quantum Computer - Quantum Computing Report - March 13th, 2025 [March 13th, 2025]
- QuamCore emerges from stealth with $9 million in Seed funding to build a 1 million qubit quantum computer - Scientific Computing World - March 13th, 2025 [March 13th, 2025]
- D-Wave says it achieved quantum supremacy using its computer - Fast Company - March 13th, 2025 [March 13th, 2025]
- D-Waves Annealing Quantum Computer Just Beat a Supercomputer Heres Why It Matters - Wall Street Pit - March 13th, 2025 [March 13th, 2025]
- Recently, a series of quantum computer-themed exchange-traded funds (ETFs) have been released in the.. - - March 13th, 2025 [March 13th, 2025]
- China unveils quantum computer thats one quadrillion times faster than existing supercomputers - Yahoo Finance UK - March 7th, 2025 [March 7th, 2025]
- China unveils quantum computer that could spell new era of processors - The Independent - March 5th, 2025 [March 5th, 2025]
- Startup PsiQuantum says it is making millions of quantum computing chips - Reuters - March 1st, 2025 [March 1st, 2025]
- A quantum computing startup says it is already making millions of light-powered chips - The Conversation - March 1st, 2025 [March 1st, 2025]
- Quantum Breakthrough: Microsoft and Purdue Unlock the Future of Topological Qubits - SciTechDaily - March 1st, 2025 [March 1st, 2025]
- Interested in Quantum Computing Investing? Here Are 4 Fantastic Picks to Maximize Your Odds of Picking a Winner - The Motley Fool - March 1st, 2025 [March 1st, 2025]
- If I Could Only Buy 1 Quantum Computing Stock, This Would Be It - The Motley Fool - March 1st, 2025 [March 1st, 2025]
- Amazon unveils quantum chip, aiming to shave years off development time - Reuters - March 1st, 2025 [March 1st, 2025]
- Quantum Computing Is Finally Here. But What Is It? - Bloomberg - March 1st, 2025 [March 1st, 2025]
- Microsoft makes quantum computing breakthrough - Drexel University The Triangle Online - March 1st, 2025 [March 1st, 2025]
- Google, Microsoft, and now Amazon: The quantum computing race is heating up - Quartz - March 1st, 2025 [March 1st, 2025]
- Groundbreaking qubit technology reduces errors in quantum computing - The Brighter Side of News - March 1st, 2025 [March 1st, 2025]
- Fortanix Tackles Quantum Computing Threats With New Algorithms - Dark Reading - March 1st, 2025 [March 1st, 2025]