Machine learning in a data-limited regime: Augmenting experiments with synthetic data uncovers order in crumpled sheets

Hoffmann, Jordan; Bar-Sinai, Yohai; Lee, Lisa M.; Andrejevic, Jovana; Mishra, Shruti; Rubinstein, Shmuel M.; Rycroft, Chris H.

doi:10.1126/sciadv.aau6792

Cited by 63 publications

(48 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, the majority of machine-learning methods require data sets that are orders of magnitude larger than those gathered in "omics" studies with limited number of patients. This is why we decided to augment our data set with computer-generated and artificially noised data to train different deep learning algorithms [48,51]. Using this approach with bile lipidomic data we selected two sets of features, lipid species, that when analyzed with NN allowed a very good separation between control patients and those with CCA or PDAC-related strictures.…”

Section: Discussionmentioning

confidence: 99%

“…On the other hand, the data structure has to be maintained. Biological data is full of correlated variables and it is important to maintain that relationship [51]. Once the synthetic data was generated as described in Materials and Methods, we applied three different reduction approaches for feature selection: DAPC, random forest (RF) and AUC analyses.…”

Section: Application Of Machine-learning Methods To Metabolomic Data mentioning

confidence: 99%

“…In addition to the intrinsic biological complexity, lack of big cohorts due to limitations in sample gathering or high analytical costs, and intragroup variability of measurements further complicate these studies. In this context, recent works have shown that implementation of artificial intelligence approaches can help to unravel disease-specific markers and pathological mechanisms even in data-limited regimes [48][49][50][51][52]. Therefore, using a novel approach, we have combined metabolomic and proteomic measurements with machine intelligence modeling and synthetic data generation [51,53,54] to identify molecular patterns that can discriminate malignant from benign biliary strictures.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Pilot Multi-Omic Analysis of Human Bile from Benign and Malignant Biliary Strictures: A Machine-Learning Approach

Urmán

Herranz

Uriarte

et al. 2020

Cancers

View full text Add to dashboard Cite

Cholangiocarcinoma (CCA) and pancreatic adenocarcinoma (PDAC) may lead to the development of extrahepatic obstructive cholestasis. However, biliary stenoses can also be caused by benign conditions, and the identification of their etiology still remains a clinical challenge. We performed metabolomic and proteomic analyses of bile from patients with benign (n = 36) and malignant conditions, CCA (n = 36) or PDAC (n = 57), undergoing endoscopic retrograde cholangiopancreatography with the aim of characterizing bile composition in biliopancreatic disease and identifying biomarkers for the differential diagnosis of biliary strictures. Comprehensive analyses of lipids, bile acids and small molecules were carried out using mass spectrometry (MS) and nuclear magnetic resonance spectroscopy (1H-NMR) in all patients. MS analysis of bile proteome was performed in five patients per group. We implemented artificial intelligence tools for the selection of biomarkers and algorithms with predictive capacity. Our machine-learning pipeline included the generation of synthetic data with properties of real data, the selection of potential biomarkers (metabolites or proteins) and their analysis with neural networks (NN). Selected biomarkers were then validated with real data. We identified panels of lipids (n = 10) and proteins (n = 5) that when analyzed with NN algorithms discriminated between patients with and without cancer with an unprecedented accuracy.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Application Of Machine-learning Methods To Metabolomic Data mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Pilot Multi-Omic Analysis of Human Bile from Benign and Malignant Biliary Strictures: A Machine-Learning Approach

Urmán

Herranz

Uriarte

et al. 2020

Cancers

View full text Add to dashboard Cite

show abstract

“…This result stems from the random training order of a randomly selected small dataset, e.g., 60 or 300 examples, consisting of a balanced appearance for each label. Around equalized trained label appearances, there are more temporal fluctuations for a dataset involving 300 examples with 30 appearances for conclusions Based on increased η with coherent consecutive gradients, the brain-inspired accelerated-learning mechanism outperforms existing common ML strategies for small sets of training examples 20 . Consistent results occur across various cost functions, e.g., square cost-function, however, with a relatively diminished performance (Fig.…”

mentioning

confidence: 95%

Brain experiments imply adaptation mechanisms which outperform common AI learning algorithms

Sardi

Vardi

Meir

et al. 2020

Sci Rep

View full text Add to dashboard Cite

Attempting to imitate the brain's functionalities, researchers have bridged between neuroscience and artificial intelligence for decades; however, experimental neuroscience has not directly advanced the field of machine learning (ML). Here, using neuronal cultures, we demonstrate that increased training frequency accelerates the neuronal adaptation processes. This mechanism was implemented on artificial neural networks, where a local learning step-size increases for coherent consecutive learning steps, and tested on a simple dataset of handwritten digits, MNIST. Based on our on-line learning results with a few handwriting examples, success rates for brain-inspired algorithms substantially outperform the commonly used ML algorithms. We speculate this emerging bridge from slow brain function to ML will promote ultrafast decision making under limited examples, which is the reality in many aspects of human activity, robotic control, and network optimization. IntroductionMachine learning is based on Donald Hebb's pioneering work; seventy years ago, he suggested that learning occurs in the brain through synaptic (link) strength modifications (1). A synaptic strength modification typically lasts tens of minutes (2) while the clock speed of a neuron (node) ranges around one second (3). Although the brain is comparatively slow, its computational capabilities outperform typical state-of-the-art artificial intelligence algorithms. Following this speed/capability paradox, we experimentally derive accelerated learning mechanisms based on small datasets, where their utilization on gigahertz processors is expected to lead to ultrafast decision making.Unlike modern computers, a well-defined global clock does not govern brain dynamics; instead, they are a function of relative event timing (e.g., stimulations and evoked spikes) (4).According to neuronal computational, using decaying input summation via its ramified dendritic trees, each neuron sums the asynchronous incoming electrical signals and generates a short electrical pulse (spike) when its threshold is reached. For each neuron, synaptic strength is slowly modified based on the relative timing of inputs from other synapses; if a signal is induced from a synapse without generating a spike, its associated strength is modified based on the relative timing to adjacent spikes from other synapses on the same neuron (5).Recently it was experimentally demonstrated that each neuron functions as a collection of independent threshold units (6). After signals arrive via one of the dendritic trees, each threshold unit is activated. Additionally, a new type of adaptive rule was experimentally observed based on dendritic signal arrival timing (7), which is similar to the slow adaptation mechanism currently attributed to synapses (links). This dendritic adaptation occurs on a faster timescale: it requires approximately five minutes, while synaptic modification requires tens of minutes or more. ResultsIn this study, dendritic adaptation was experimentally examined at a higher stimulation fr...

show abstract

“…The importance of the descriptor selection from physical considerations has been observed for a diverse set of materials science applications; [1][2][3][4][5][16][17][18][19][20][21][22][23][24][25][26] however, it is not always possible to find the relevant physical descriptors for the desired application. Furthermore, even if physical descriptors have been identified, they are not always easily accessible.…”

Section: Article Scitationorg/journal/jcpmentioning

confidence: 99%

Screening billions of candidates for solid lithium-ion conductors: A transfer learning approach for small data

Cubuk

Sendek

Reed

2019

The Journal of Chemical Physics

117

101

View full text Add to dashboard Cite

Machine learning (ML) methods have the potential to revolutionize materials design, due to their ability to screen materials efficiently. Unlike other popular applications such as image recognition or language processing, large volumes of data are not available for materials design applications. Here, we first show that a standard learning approach using generic descriptors does not work for small data, unless it is guided by insights from physical equations. We then propose a novel method for transferring such physical insights onto more generic descriptors, allowing us to screen billions of unknown compositions for Li-ion conductivity, a scale which was previously unfeasible. This is accomplished by using the accurate model trained with physical insights to create a large database, on which we train a new ML model using the generic descriptors. Unlike previous applications of ML, this approach allows us to screen materials which have not necessarily been tested before (i.e., not on ICSD or Materials Project). Our method can be applied to any materials design application where a small amount of data is available, combined with high details of physical understanding.

show abstract

Machine learning in a data-limited regime: Augmenting experiments with synthetic data uncovers order in crumpled sheets

Cited by 63 publications

References 45 publications

Pilot Multi-Omic Analysis of Human Bile from Benign and Malignant Biliary Strictures: A Machine-Learning Approach

Pilot Multi-Omic Analysis of Human Bile from Benign and Malignant Biliary Strictures: A Machine-Learning Approach

Brain experiments imply adaptation mechanisms which outperform common AI learning algorithms

Screening billions of candidates for solid lithium-ion conductors: A transfer learning approach for small data

Contact Info

Product

Resources

About