108 research outputs found
105 - Alex M Fout
No full textIncludes bibliographical references.This poster was presented at the 2017 Colorado State University Graduate Student Showcase, 9 November 2017.Determining the interface between two interacting proteins can help illuminate cellular biology, improve our understanding of disease, and aid pharmaceutical research. Such determination is expensive and time consuming using wet-lab experiments, which has motivated the development of computational methods. Inspired by the success of deep learning in image processing and other application areas, we adapt convolutional neural networks to work with irregularly structured data, such as proteins. We construct a novel pairwise classification architecture which is trained and tested with data from the Docking Benchmark Dataset versions 4.0 and 5.0. This outperforms the existing state-of-the-art prediction method, PAIRpred.Graduate Student Council - New Graduate Student - Research Top Scholar
Persoonlijke fout bij astronomische metingen
No full textBehandeld wordt de persoonlijke fout bij astronomische metingen. De nadruk wordt gelegd op het onderzoek van twee instrumenten, een motor-micrometer voor de Wild en een voorzetprisma voor de Zeis Ni-2A (evenals bij de Wild T4, voorzien van een stappenmotor), beide ontwikkeld op het Lab. voor Geodesie in Delft. Voor het onderzoek zijn kunststerwaarnemingen (alleen met de Wild T4) en veldwaarnemingen (op de grote pijler van het Lab. voor Geodesie en op de oostelijke pijler van het satelliet-station in Kootwijk) gedaan.Civil Engineering and Geoscience
De fout vriendelijkeorganisatie...
No full textNearly every organization is faced with the consequences of human error. This report describes how a newly designed method can be derived from the analysis of errors in handling in the construction process of the Dutch building industry. What has not been examined in this research is any technical solution with respect to how the building industry can avoid errors or how one can prevent these errors from occurring. This research is based on the theory that, if an organization is able to change the way it deals with mistakes in a positive way, the chance that people will speak up about their errors will increase. This can provide an organisation with the ability to prevent potential negative consequences due to error in the early phase of the building process. What appears in this research is that a culture of blame develops and the feeling of shame after making a mistake increases; this can be an obstacle to learning from one’s mistakes. Another aspect of the culture of blame is that people do not speak up, or they conceal their mistakes which then can lead to negative consequences due to error escalating because no one noticed.DCM, Building Process innovationReal Estate & HousingArchitectur
Een weg terug om vooruit te komen: Lars Spuybroeks The Sympathy of Things
No full text"Als de heer [John] Ruskin gelijk heeft," schreef een recensent van zijn boek The Stones of Venice kort na de publicatie in 1853, "dan zitten alle architecten fout, is het hele architectuuronderwijs van de afgelopen drie eeuwen fout geweest." Ruskin reageerde daarop in een volgende uitgave van het boek: "Dat is precies wat ik heb willen zeggen!" Lars Spuybroek wil in zijn publicatie The Sympathy of Things: Ruskin and the Ecology of Design hetzelfde beogen.ArchitectureArchitecture and The Built Environmen
Een data reducerende codering voor "schrijfsignalen”
No full textVoor broncodering van "schrijfsignalen" worden enige digitale coderingsmogelijkheden, P.C.M., D.P.C.M., A.D.P.C.M., vergeleken. Onderzocht wordt welke voorspelling van de plaats van de pen op het schrijftableau uit voorafgaande plaatsen van de pen, de gemiddelde kleinste fout in de voorspelling maakt en welke codering deze voorspellingsfout met zo weinig mogelijk binaire symbolen per tijdseenheid beschrijft. Gebleken is dat bij een codering, waarbij afhankelijk van de grootte van de fout D.P.C.M. of P.C.M. gecodeerd wordt (dus eigenlijk A.D.P.C.M.) een minimum aantal binaire symbolen per tijdseenheid nodig zijn om de plaats van de pen bij de ontvanger exact te reconstrueren.Electrical Engineering, Mathematics and Computer ScienceTransmissie van Informati
NLINLS: a Differential Evolution based nonlinear least squares Fortran 77 program
No full textThis paper provides the list of Fortran 77 codes of nonlinear least squares using Differential Evolution as the minimizer algorithm. It has been tested on a number of difficult nonlinear least squares problems (taken from NIST, USA including CPC-X Software challenge problems). Help on how to use the program also is provided.Nonlinear least squares; Differential Evolution; Fortran 77
Supplementary Data for NIPS Publication: Protein Interface Prediction using Graph Convolutional Networks.
No full text<p>These data sets can be used to re-run the experiments from our paper, Protein Interface Prediction using Graph Convolutional Networks. The data are derived from protein complexes in the docking benchmark dataset v. 5.0. Each file is a python tuple that has been saved using cPickle and compressed using gzip.</p>
<p>Links:</p>
<p>Paper: https://papers.nips.cc/paper/7231-protein-interface-prediction-using-graph-convolutional-networks</p>
<p>Poster: https://zenodo.org/record/1134154</p>
<p>Code: https://github.com/fouticus/pipgcn</p>
<p> </p>
<p><strong>File Descriptions:</strong></p>
<p>train.cpkl.gz and test.cpkl.gz have the data formatted for neighborhood based graph convolutions. The diffc_ files are the same data formatted for the diffusion convolutional neural networks that we compare against. </p>
<p>train.cpkl.gz is a tuple of length 2:</p>
<ul>
<li>element 0 is a list of length 175 containing the PDB codes from the docking benchmark dataset</li>
<li>element 1 is a list of length 175 containing features for each protein. Each element is a dictionary containing the following keys:
<ul>
<li>r_vertex: vertex (residue) features for the receptor. numpy array of shape (x, 70) where x is the number of residues in the receptor and 70 is the number of features.</li>
<li>l_vertex: vertex (residue) features for the ligand. analogous to above, with shape (y, 70) where y is the number of residues in the ligand.</li>
<li>complex_code: PDB code of the complex. matches the list of codes described above.</li>
<li>l_edge: edge features for the neighborhood around each residue in the ligand. numpy array of shape (y, 20, 2) where y is defined as above. the second dimension is the edges to the 20 nearest neighboring residues, ordered by decreasing distance. The third dimension allows for two features per edge. </li>
<li>r_edge: edge features for the neighborhood around each residue in the receptor. numpy array of shape (x, 20, 2) where x is as above. </li>
<li>l_hood_indices: the index of the 20 closest residues to each residue, ordered by decreasing distance. numpy array of shape (y, 20, 1). "Index" means which row in l_vertex gives the vertex features for the closest neighbor, second closest neighbor, etc. </li>
<li>r_hood_indices: analogous to above, shape (x, 20, 1).</li>
<li>label: 1 or -1 label for each residue pair. numpy array of shape (x*y, 3). Each row looks like (i, j, k) where i is the index of the ligand residue, j is the index of the receptor residue, and k is either -1 (negative example) or 1 (positive example).</li>
</ul>
</li>
</ul>
<p>test.cpkl.gz matches the structure of train.cpkl.gz except it has the test set of 55 complexes. </p>
<p>Descriptions of the vertex and edge features can be found in Appendix A of <a href="https://mountainscholar.org/handle/10217/185661">this.</a></p>
<p>diffc_g2_p2_train.cpkl.gz is a tuple of length 2:</p>
<ul>
<li>element 0 is a list of the same 175 PDB codes as above. </li>
<li>element 1 is a list of features for the 175 complexes. Each element is a dictionary of features with these keys:
<ul>
<li>r_vertex, l_vertex, complex_code, label: these are the same as described above. </li>
<li>'r_power_series': Stacked diffusion matrices which are powers of the similarity matrix used in the DCNN method. numpy array of shape (x, 2, x) where x is the number of receptor residues. the middle dimension 2 indicates how many "hops" is used for that diffusion (1 vs. 2). In other words, element (i, 0, j) is the similarity after 1 hops between residues i and j. element (i, 1, j) is the similarity after 2 hops. See DCNN paper for details.</li>
<li>'l_power_series': same as above but for the ligand. shape is (y, 2, y).</li>
</ul>
</li>
</ul>
<p>diffc_g2_p2_test.cpkl.gz is the same as diffc_g2_p2_train.cpkl.gz but for the 55 test complexes.</p>
<p>diff_g2_p5_train.cpkl.gz and diff_g2_p5_test.cpkl.gz are the same as the p2 version above, except that the diffusion matrices have shape (x, 5, x) and (y, 5, y) because one of our comparisons against the DCNN model uses 5 hops instead of just 2. </p>
<p> </p>
<p>Note: these files were pickled with Python 2.7. If you're unpickling with Python 3.x you might have to specify encoding as 'latin1'. </p>
<p> </p>
<p>Please direct any questions to:</p>
<ul>
<li>Alex Fout ([email protected])</li>
<li>Jonathon Byrd ([email protected])</li>
<li>Basir Shariat ([email protected]</li>
<li>Asa Ben-Hur ([email protected])</li>
</ul>This work was supported by the National Science Foundation under grant no DBI-156484
Enhancement of the Bessel computation in the COS method
No full textIn dit verslag bekijken we het prijzen van Bermuda opties met de COS methode. In deze methode vinden we de Bessel functie waarvoor de rekentijd lang is. Daarom bekijken we een benadering die sneller is en de orde van de fout niet verandert.Applied mathematicsElectrical Engineering, Mathematics and Computer Scienc
- …
