Machine Learning at Scale: Astrophysics

Josh Bloom UC Berkeley (Astronomy), LBNL @profjsb ML at Scale:
Astrophysics AI for Science Town Hall 2019 Sept 11, LBNL Data Driven Discovery Investigator

E. Ramirez-Ruiz (UCSC) (cf. Kasen) 50 100 150 200 Days
Since Explosion Type Ia NS + NS Mergers Type IIp NS + RSG Collision IMBH + WD Collision Pair Production Supernovae -10 -12 -14 -16 -18 -20 -22 M H z=0.45 200Mpc Rumsfelian Challenge LSST 20 Tb/night (raw imaging) 150 TFLOPS (First Data Release ~ 2022) ML Impact at Scale Optimizing Payout over known-knowns, known- unknowns, & unknown- unknowns (Small # of labels of known knowns) ie. Discovery & Classification To Prioritize Followup with Scarce resources

Supernova Discovery in the Pinwheel Galaxy 11 hr after explosion
nearest SN Ia in >3 decades ML-assisted “real-bogus” discovery at LBNL ©Peter Nugent Nugent, …, JSB+12 1110.6201

Discovery (& classification) on images is now a cottage industry
While this is ML at scale, it is hardly taxing computationally Adapted from D. Goldstein

Cosmological Parameters from the Dark matter Distribution Figure 6. The
architecture of our 3D conv-net. The model has six convolutional and 3 fully connected layers. The first two convolutional layers are followed by average pooling. All layers, except the final layer, use leaky rectified linear units, and all the convolutional layers use batch-normalization (b.n.). Figure 7. (top) visualization of inputs that maximize the activation of 7/1024 units (corresponding to seven rows) at the first fully connected layer. In this figure, we have unwrapped the maximiz- ing input sub-cubes for better visualization. (bottom) magnified portion of the top row. vation of a particular unit while treating the input X as the optimization variable (Erhan et al., 2009; Simonyan et al., 2013) X⇤ = arg max X s.t. Xl,i kXk2  ⇣ “Estimating Cosmological Parameters from the Dark Matter Distribution” Ravanbakhsh+ 1711.02033 “Fast Automated Analysis of Strong Gravitational Lenses with Convolutional Neural Networks” Figure 1: Comparison of parameters estimated u true values (x-axis). From left to right, the panels y components of complex ellipticity. The shade Hezaveh, Levasseur, Marshall 1708.08842 Black-Box Cosmological Inference Einstein Radius (model) Einstein Radius (CNN)

Challenge & Opportunity: Domain Adaptation Figure 1: Diagram of an
RNN encoder/decoder architecture for irregularly sampled time series data. This network uses two RNN layers (specifically, bidirectional gated recurrent units (GRU) [6, 25]) of size 64 for encoding and two for decoding, with a feature embedding size of 8. The encoder takes as inputs the measurement values as well the sampling times (more specifically, the differences between sampling times); the sequence is processed by a hidden recurrent layer to produce a new sequence, which can then be used as the input to another hidden recurrent layer, etc. The fixed-length embedding is constructed by passing the output of the last recurrent layer into a single fully-connected layer with linear activation function and the desired output size. The decoder first repeats the fixed-length embedding nT times, where nT is the length of the desired output sequence, and then appends the sampling time differences to the corresponding elements of the resulting vector sequence. The sampling times are passed to both the encoder and decoder; the feature vector characterizes the functional form of the signal, e.g. Unsupervised feature learning using stacked RNN autoencoder for irregularly sampled time- series, using measurement uncertainty in the loss. Naul, JSB, Perez, van der Walt (2018) 1711.10609 •Architectures and platforms are designed for well-measured images, video, graphs, & text. Our data are different. •But…Our metrics tie directly to inference, with physical meaning •“Small label problem” - expensive to obtain/simulate training data & labels. Need to exploit self-supervised, semi- supervised, and transfer learning.

ML for Novel Discovery & Inference

https://www.forbes.com/sites/jillianscudder/2017/01/21/astroquizzical-care-planet-nine Searching for Planet 9 hypothesized to exist using mostly
public data about comet orbits + deep domain knowledge + sophisticated computing Batygin & Brown 1601.05438 A Modest Prediction: discovery data of Planet IX has already been obtained & resides in existing public data archives. It will be a group of clever astronomers & statisticians with a lot of compute resources that will make the retrospective discovery of the century… Massive new planet beyond the orbit of Pluto

M. Medford, Nugent, JSB et al. •More candidates than disk
space at NERSC •Needed a fast (~10 ms prediction) ML pipeline to decide what data to save •Optimize scoring wrt predict speed, distributability Today’s Compute Architecture Searching for Planet 9 Method: shift & add images by allowable orbital phase space Data: 103k Palomar Transient Factory (PTF) images spanning 795 deg2 over 7.5 years Can distributed model servers efficiently access massive (shared) GPU resources?

Faster-than-GPU Inference is Already Being Explored in Particle Physics 1904.08986
See also "Quantum-enhanced machine learning” in Carleo+ 1903.10563

See also Lang, Hogg, Schlegel 1410.7397, Leistedt+ 1807.01391 Forward-Folded Cosmology
Inference Figure 2: The Celeste graphical model. Shaded vertices represent observed random variables. Empty vertices represent latent random variables. Black dots represent constants. Constants denoted by uppercase Greek characters are ﬁxed throughout our procedure. Constants denoted by lowercase Greek letters are inferred, along with the posterior distribu- “Learning an Astronomical Catalog of the Visible Universe through Scalable Bayesian Inference” Regier+ 1611.03404 LLNL-PRES-733055 Our hierarchical Bayesian forward models c tolerances for galaxy shear when the mode The forward model of our galaxy image data “Hierarchical probabilistic inference of cosmic shear” Schneider+ 1411.2608

Inverse Problems & Likelihood-Free Inference (a) DRDAE: quasi-circular (b) DRDAE:
pure noise (c) DRDAE: eccentric (d) EDRDAE: quasi-circular (e) EDRDAE: pure noise (f) EDRDAE: eccentric signals Fig. 4. Performance on GW signals contaminated by real LIGO detector noise with SNRpeak = 0.5. The plots include reconstructed outputs of DRDAE and EDRDAE on quasi-circular signals, pure noise input, and eccentric gravitational waves. Table 3. Ablation study for major parts in EDRDAE MISSING PARTS W/O SA W/O BA W/O CL resilience of the model to denoise signals that are not used during training, i.e., eccentric GWs. Currently, there is no "Denoising Gravitational Waves with Enhanced Deep Recurrent Denoising Auto-Encoders” Time (sec) Shen+ 1903.03105 cf. ”Fast likelihood-free cosmology with neural density estimators and active learning” Alsing+ 1903.00007 Turn inference into density estimation task using simulated data 20 Simulation Machine Learning Inference x z <latexit sha1_base64="HjZ6RxRDdZu139wdkhmGLAXlGyY=">AAAoEHicpVpbc9vGFWbSW6q4TdI+5mVdxbHlgDQBUbacjDJuLpN2Jm4cW3bSCpJmASyJHeLmxUImhaI/otMf07dOX/sP+tbH/oe+9JxdkMSV0jj0GAR2v/Ods2fPnj0LykkCnsrx+N9vvPmjH//kpz976+c7b9/4xS/fefe9X71I40y47LkbB7H43qEpC3jEnksuA/Z9IhgNnYB958w/x/7vLphIeRwdy2XCTkM6i/iUu1RC0/l7b//NdtiMR7nk88uEuzITrDg53SHExpZULgOWu3EUMRcFiqOT1I+FZBH59MhMpCF97s4N4gn66sgJqDu/+WBskGH+6Unq0oAdmaeFQaLYY8SDwdDIZUd2FAV0CTaxpKFnmkXdWmAES7JSRVOfJFRKJqKjPI7IfiJJPJ0SK5HFdksa6iKWCaUMZQyXCzdg5YCmPAiOXvlcMiPkEQ+zkKT8UtmOg8H7BtmCrOn0TdUUg2wIG3LSZ5L2yHpUzIdOkLFSfv180xzf7CLzueeB07ZY0vQ4j2jwmpYvSEAdFhTkiJxItpDaPMG8bqMa4H5rulhngrGo03NdaHTR6Q6gbxEFQrlpDMvC9Ykd0ZCRe8ReEh6RfGyMRiPDLBCCM3tSnY9Tckc9DpXQHqGS3BkblhuSIYi74R7Ji0+UmsWVKqyNisWGflGnHpoNbkXuxAuWojROywnGP/OMcjnU4wQYh+PRfWM8muyR4RAomw/W6L5+GLafEPhJt54V8XB/JdHxtFY07HpEbDkg37zSXZONu2pBDYb4ZtVplWxiEHN0QNCDZONCUqq0fpBKq6rSultV2qFTqwzimVgzVhcaEGKfJtuvkRHw1UGNJTWIgEkg+NBJJYdjzTSpM1lrB/RLmp2S5pWSaadcxfZ+UdEtalUn63O93+jFC55txIve/FrLd90P+Vm25hV2DP9ks5FtlrZm2yPMmzEdXCi/17RlcQ07rNeyY3FtG3wTjbBG262YvJYVqLRhhrUyo2GFBVasgvsqMzoUWU1FajE0x4qNoKcMfxVOimtVI6wXkeZIVebajhHXwOBqKqN3O8xsumUVrN226mAba/kTh0UeCdhUHpnjcgl3m6bFzJaYNT6tGfFbKEsklayy7irb46lackO14uARVhyB/DCyDnTC+sBW2PPxB7WFu9giPDQ30ou6WHW717L7o3aeKFPmhlTZAS4mtk9lLooPOpJIhbOWP8hHVaa1AkWJbCmyvTaZVaX6YYaVG9SaTWo2G+a1VoCfv7s7Ho3Vh7RvzPJm9xH59r//GQwGT87fe9+1vdjNQhZJN6BpemKOE3maUyE5VLXFjp2BAVBxUQgjuMXNLD3N1fmhILegxSOwlOF/JIlqrUrkNEzTZegAMoQATZt92NjVd5LJ6eFpzqMkg1re1YqmWUBkTPAwAmcCAUsnWMINdQUHW4nrU0FdKO7rWpRNxkIbvLODyw7uUrA1SuFgAcMmr7iEg0EQS5D0GEwMU+h8VZYWuZg5RT4eHRroS7yYh+P9h5Z5cP/wvvVgcgD7SFsS66q16Hgteg1JVbTWRasy44fAofiKnVtt6VjQaLbSbGrx+5MH44PDQ2v/wHo4Mc3DUvoK4XLEkxLdNXuGl+LUG3gv4zhIaxGTpzyLuFzUG2eCJlAbNlrDLJBcxK8MJ47nkjqpAZcsoGJhTIOYynoootKjKBYhDdSZKk8zZ8pnDe2BLkUdQeesTpBP2TIKkyHNZFxnTuH8+KEbh3AaTiHSIyod7gDkCSyObxJMr+lx/KRk8ZeJz6K0yDMRFFWWxJvCscKAnIYn7Hlq4FX5+agSHQZx4XhUbdZTbxDgqzZjFCrHhPCUxgmLFKGM3SMaBKdoBxOCTetjZFEWAn8I60ctUugSQz3lHgHfBkzWFwt4vOGOnLourJF0RWH7kH7qMpB+6iKLFO4YSqibFPd2cGhII0+rQ5GAw6yIJSQamHPYrD1wg1DvFtIRDpJHM2jVvaMQdhm1eL9iEROQNzEJAI8E1I4dsVclfW5jDGJsFCfmKTzByS51812zKOowuE8gXahILHLYPQVk5QSMd6AqmdsOn6VzntTaophHsD/JBhPEJ7/AudAafZiE3Jcy+fjePdU1isXsHkTzPTBCGyQlDPp7fvExmtVgwxu2sh7cMIMUSEVup9MpDXmwtFPIdonEkFdczUSlKSuOwuwLq5LU1fCwQHrpizDnTRs8Pp1uur1mN5sVec5GtjEbFX9p9HHIHDmHPtbuYy8zfkEDHB2Mh4fspTL0jxAfUUxCunRYXWDJEkDC6oKNDY1B7zRTJBSKrs9ctSxazoQY8uJpP4dO0EBBU9kShqln/aLK2eqlAmyHpcufqWCvs7x8+TKjALXVN9FfRQuzBq1QPbANbg3USFD/xF+m3E1XM14XhgGsYkr6Lg3yx8W5CqCOtYEDr2G/Kc47YMFM1GFfb6EUXpPSxnL0zq5pCz7z5V5DwBGbCPzsaZMuTCCG6nODV2eaJ8X58ZleZnnIU5yaljBrCn95pQxsV+nLQn2f2R6dzRhkQnhowKBGgXB7dl6S4ZMOjeewCTZmhF2oZZCzF62wXXXN233hqu9xu2+26vuq3SdXfcftvqmju+C7uZw3XWf5sJ2tkrI7aUtuulaS4IYvoOAS3MlUlifN9YZ7eXFiVaLkD4V9E/6VkULsEE4IAfsz2bXIOmyQlgnILJJfQCoBLrXXZHCoiUXD54nHhNYwhVKRqLwvOWRJiNjqk9VyX1MQc2QppW/bIp0yZ1ZVKgexs5bktCXIAjiQlXL63tIO/SwOvO4FDy2eOpzVY52oDi2R69Nba1YREWVFvxh0doiwJOVBHG2Ru6BiBeqQXzQkyxW96MJedmMvu7DLbuyyC3vRjb3owspurOy0l4m4Gz7WE/mYST/2GnPoUigq10XM5/hkl7VVHSjiQGyAT/GpG5hCsbjcIJ+px34oj+I6GBu64S7FH0uq1qrnHnsb4KcNsNpMmYu/NpEYglR0x7gKXyx6y9tV3sWmVkKKJJ4OqRM0qgqSdxZSyTYGcR0GsY0hvQ5Duo1BXoehXdZUGC6vw/CnFkMQw1yFMSSia/lxNSlKTG+tML2/jySD4rs5o2AdpgH4IvZNopPkwv6kPYZFIqDyakFvd2J5C3fOO4HzNnDeBZzStpk6hyG6Bb9sYi97SFtAnTw6bW1hF8Q2qvwtESimGGvafbbf7eBpnIkWdtKPXbaxyz5s03TAdnsEbUihKgi6DDH6hlidxerWu0/msIPeseyE753tF30a+8QnVfFJp7hSn/SqT66jvk98UhXvVq/Ee/yVkHtkJaxcd4u8YC6WSKpkghUseOv0Il/FF8zVtYj+sT9RwMXd4sQ9JViS2aoYw9eQ665mIYROUTT7V9HAdX8rF45QUU2uQ6Wuk+3WlYyim1Jck7LlNnXe6jfyQyDUfB9eaaHyn+J7uJUPL/vaPLg7wMt9Ze4DvD3Ey8PtipTZWz3xOnaLbsPFDzIcwvfrOFLHMIeKxmHzggloLA+ccyYifHseZqoD/9ylbB2qVjiBVp6KqoA+XMQZ7JVEvfUCLQGLZnjexNdgPsMzSGPY2KFU39oh8LHxLVWskTWp+vEV3+mrdzelFPU8GXfoyofj0f4BZPESh3L2ReLTSMYhFFZZwHITT8dVoRUYUkNa1aUrT9y74cgvyvPSF8wNoKjC5m/KVphLqHjgf0/vMfQe9/Z6nM6KXF17EE8ZFoNPWV//k5hDvaSuPQg3hsodLz39cPIocrz09LMFlJtSvX0s3zs4Tv5l0Qf/7GmBbyV6en23yP2RbbijbsRdfGk4Cylsr/BtG3i3DcijFRDuenQKFmTowq+f9yGCeAanbw62re/6lD5+9mWR46UPIEXIKFglQxVCx3x+qX5KqPyRUz4ePThww2LVvvphK7dqjfr3rSbYiYWn0eoPAyrti9TnUwnNika3B2lFqbXBl+1Mb4moo92FP3G0W2HIM1hHlY61+a1G/Sa6Oi7saIzr/N1ds/kjXPvmhTUyxyPz2/Huo0cD/Xlr8P7gN4M7A3PwYPBo8LvBk8Hzgfv2/27cvHH3xke3/3r777f/cfufGvrmG6XMrwe1z+1//R+6gFM9</latexit> <latexit sha1_base64="N0otbztN+sjGy9xqdVKMRIGUlNA=">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</latexit> <latexit sha1_base64="N0otbztN+sjGy9xqdVKMRIGUlNA=">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</latexit> <latexit sha1_base64="KGKYmyKWjUVupNzZY8DVUShrvqU=">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</latexit> r(x, z|✓) <latexit sha1_base64="n4ECwTYdCfuLwJXCZl6+ybiKm1E=">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</latexit> <latexit sha1_base64="yfNfOs7ZyWqkNT8RpYmkj2bn/mA=">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</latexit> <latexit sha1_base64="yfNfOs7ZyWqkNT8RpYmkj2bn/mA=">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</latexit> <latexit sha1_base64="kQLQGJ4Xe+WRfsBc6Tu8qUqoqpg=">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</latexit> t(x, z|✓) <latexit sha1_base64="BMHppIHEY/0iNNG8Fhs+Hd4wokw=">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</latexit> <latexit sha1_base64="WhfdL6HX9XmJ7WXSrLsNwv05onI=">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</latexit> <latexit sha1_base64="WhfdL6HX9XmJ7WXSrLsNwv05onI=">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</latexit> <latexit sha1_base64="GuTW9do3dzdd+I9ZbAmjhvNlqcg=">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</latexit> ✓ <latexit sha1_base64="uKih3q81UG5Y39UkSUDmaMgSwo0=">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</latexit> <latexit sha1_base64="sg4BK8w5ytAcXmQyQ0tWWWn7/Y0=">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</latexit> <latexit sha1_base64="sg4BK8w5ytAcXmQyQ0tWWWn7/Y0=">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</latexit> <latexit sha1_base64="cLB/gHcLXKsvdN1Ss3kTuhimZm0=">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</latexit> arg min g L[g] <latexit sha1_base64="K5zdKfNQlfWV+Iu2lZ+kbdAR2qQ=">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</latexit> <latexit sha1_base64="ApOBYqsfgOriMeFGC7CAo7I3urI=">AAAcqHicpVlfc9y2Eb8kbZMq/eOkj/YDU41cx3M6350kW07HMx4nnrQzdq3Kku1UlDQguSQxBxIUAJ7vxLKfoG/9NH1tP0W/Rp/70AXIk/j3rGlPIx6I/f0Wi8ViueA5CaNSjcf/+ujjT3704598+tlPNz7/2c9/8ctbX3z5RvJUuHDscsbFO4dIYDSGY0UVg3eJABI5DN46s2+1/O0chKQ8PlLLBE4jEsTUpy5R2HV+a3vLVrBQRlHmETHbdlgKeWarEBTJN2wigojG54H14iQ4Pb+1OR6NzcdqNyZlY/Pp7X8/+/Tzv/7n4PyL267tcTeNIFYuI1KeTMaJOs2IUNRlgPpTCQlxZySAE2zGJAJ5mhlzcmsLezzL5wL/Y2WZ3iojI5GUy8hBZERUKJsy3dklO0mVv3+a0ThJFcRuMZCfMktxSzvJ8qgAV7ElNogrKNpquSERxFXoytooxqbhojB4Y2PLMi6XaGsscfVw2tZ7qkIrYVwh0wMfF6ribQFenonAybPxaH+ofakvk/3xzuPpZO/h/sPpo929Sd7BLNappI6vqDdgBgIgrlOrnPFj1GH05RtbbTYXJA5WI08K+sPdR+O9/f3pzt708e5ksl+yP0AuZ7xbortWb+hJvfRD3VacM1mLmEzSNKZqUe8MBElC6jZ6o5QpKvj7ocP5TBFHDvGSMiIWQ59xouqhqAd9EnMRESbpJZxmMnV8GjRGx4AOwRs6gsygriDzYRlHyTZJFa9rllyouy6PcJdKjPSYKIc6CDnAzfEq0ZtSHvGDUku4TEKIZZ6lguVVLYnn47YdhtTTO38mh/pq/PykEh1Dy6UKqt3F0g8t1Fft1lFoHBPhneQJxEah4u4TwtiptgOEAL8+R4jTCPVHuH/MJkWR2C6W3LPQtwxUfbOgxxvuyIjr4h6RKxV2GBJV59DZZZ2ykNgCzTANadEYt1sUkdgrhtMURnFVxBITDa65HFoeukGYnCdHepI0DrC3kI4ik+pw834PMQjCTBJAPQpRG3YM70v1mBUxBnVs5CeT08xkTulmm5M8r8OwnWC6MJGIuRTbdswTNN7B3DyzHRrIGU1qfTGnsYeuaGjC+KRzvRbFiCEuQhYqlXzz4IERjbgIHmA0P0AjCoOUwkm/o/NvtFkNbboBK+vRDQGmQCIyW/o+iShb2hKzXaJ0yDcfCyZRFSorjtLZF3elVR+GRrlWr0IRZbRpg0d9/1rsNcUQ5FkGI3sYjPK/NGQUM0dGUQZtGVykdE6Ynh3Oh0ZwYQz9AeMj5lZElg7UCUtIEIm7KxWgjdHeaaZIy7LdEFyzLVrOxBjyuN+vo0jQqIJI1SLj0kM/1TgbmQrnIUuXvzbBXtdycXGREoTa5tsqvvIW5gq0QvXArnFXwAKJwx+ES0lduVrxOhknsIopFbqEZS/zcxNAHXtDT7yGfZWfd8BYIOqwF2tUCq+p0mbgq3ubE1vQIFRfNwiOuI7AZ4dNdVGCMVRfG311/CzJz4/Oim2WRVTqpWmRoUl+/kEOPq7kRW6+z2yPBAFgJsSbBgxrFAy31+elMn1XhMYxPgQbKwJzsw0yeNMK25Vo1pZFK9nLtixYyb5vy9RKdtSW+U4hwu/mdr4WnWXb7WyVlOKkzbwWrZjohu+w4BLUSU2Wt5r7TT/L85NpJUr+kNtf4V8ZKZYdUc9j8Gdrc2pdhY1WCwIzi6JzTCWoyzxrUkUUFw2fJx6IYgQfS0XL5H1FMUtixFbvpi33NYk6R5asotmmdHLOplVWhrSzFtNvEYEpsuIV7Wnh0Geced0bHnu84oRQi3XLCArG6gjRHF8j4jTvp6GwgwKJpIzHa3hzIlagDv6iwSx39KILe9mNvezCLruxyy7svBs778KqbqzqtBcE74aPi4V8CSrkXmMNXYJF5VUR862+s8vaqg4UnIlr4KG+6wZKLBaX18jX5rYfSmNeB+uObrhLJO7YqrXmvsfeBviwATYPU3D1KdjiGKSiO8ZN+Oqit2yu8q7uaiWkWOnTIcFjd72qsLLOQipZp0HcRINYp0HeRINcp0HdREO7rKlouLyJhj+1NDCOaxVxTEQ38uNqUQyteLTi8v4+VoDFd3NF0TqdBvDLsr+yiiS5sH/bnsMiEVh5taC/6cTSFu6cdgJnbeCsC+iTtplFDtPoFvyyib3sUdoCFsmj09YWdmHZw6r+FgWLKYCm3Wc73Q72eSpa2N1+7LKNXfZhm6Yjttsj2gaJVQHrMmTYN8XqKlYfvTvWDJ+g96Z2Qr8+28n7Ruyj71bpu510M3zSO3xyk+H76LtVevfwht7jr8R6YK3IxnVb1htwdYlkSibcwYK2Ti/qPZ+DW9QiDuBZNEsMcHE/P3FPLV2S2aYYA0RfiZqFkHaKUbPzITV43VmrS8/QqNq9iSpz3V1vXalRdKsUN1TZcps5b/UbeRcVFvruftBC4z+j7/FaffqyU5iHrT19eWjMfaSb+/ryeP1Axuy1nvhf7Bbdhov/y3AM3xc8Nscwh4jGYXMOAjvLA+cMRGxNRntRagT6NXzZu2168QRaucurhOJwwVN8VlrmrReOwiAO9HlTvwYLQZ9BGtPWAjP01oaFH1u/peIFssaqH1+xv3h3U7KI5yneMVa2PR7t7GEWL3GaZ8+TkMSKR1hYpQyyiT4dV0krMKYGWR2rqDz1sxuP/KI8L30HLsOiSne/KntxLbHiwf8e6RFKj3qlHiVBnplrD+IQdDF4CH3yA06xXjLXHoTLsXLXlx45njzyTF965LDAclOZt4/lewfHyZ7nffBnh7l+K9EjDd08C0f20B11I+5n5jcbgo9X/LaHurUOSOMVEFs9YwpgqXbhi+M+BOMBnr4p2nbV6hv05evneaYvfQAlIiBolYpMCB3R2aX5KcGO4xgLQ/12MhuPHu25Ub7qZ2QJQkKSTWudDjDd2QA7XHgFejyq9y9kSH2F3UZN0c9kZdDpNb7sh+KRqMdoi/RPHO1enHKA+6giuDK/1Vm8ia7OSwsa8zq/tTlp/gjXbryZjibj0eSP482n+4Pi89ng9uDXg3uDyeDR4Ongd4ODwfHAHfxt8PfBPwb/vHP/zsGdt3d+KKAff1RyfjWofe44/wW/t7R6</latexit> <latexit sha1_base64="ApOBYqsfgOriMeFGC7CAo7I3urI=">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</latexit> <latexit sha1_base64="Y+642ZFyiyZWQmFYdFZi+qnOJts=">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</latexit> ˆ r(x|✓) <latexit sha1_base64="NPA37tSEX6NJZFS7lgnMywc24Uw=">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</latexit> <latexit sha1_base64="OWtH3yFwQxlgBBM8LzzB+Kr7/wE=">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</latexit> <latexit sha1_base64="OWtH3yFwQxlgBBM8LzzB+Kr7/wE=">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</latexit> <latexit sha1_base64="8WBmztmqRcTSWIh+NwhKhjUiPqA=">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</latexit> ✓i <latexit sha1_base64="l7N+1zwpgVSD9LHyu2N3Zdptxos=">AAAcgXicpVltc9y2Eb4kfUnVpnXaj/EHphq7duZ0ujudbDkZz3iceNLO2LUqS45bUdKA5JLEHEhQAHi+E8v+iH5tf03/Rf9NFyBP4utZk55GPBD7PIvFYrFc8JyEUanG4/9+9PEnP/npz37+6S+2fvmrz379mzuf//at5Klw4cTljIt3DpHAaAwniioG7xIBJHIY/ODMv9XyHxYgJOXxsVolcBaRIKY+dYnCrne2CkGRC3pxZ3s8GpuP1W5Mysb2s8/+k1qDweDw4vMvXNvjbhpBrFxGpDydjBN1lhGhqMsg37JTCQlx5ySAU2zGJAJ5lhmDc+se9niWzwX+x8oyvVVGRiIpV5GDyIioUDZlurNLdpoq/+Aso3GSKojdYiA/ZZbilp695VEBrmIrbBBXULTVckMiiKvQR7VRjE3DZWHw1tY9y/hSoq2xxGXBaVvvqQqthHGFTA98XAGDzjwi5jsCvDwTgZNn49HBUPtSXyYH470n08n+o4NH08ez/UnewXRYCtfU8TX1FsxAAMR1apUzfoI6jL58616bzQWJg/XIk4L+aPZ4vH9wMN3bnz6ZTSYHJfsD5HLGsxLdtXpDT+qlH+q24pzJWsRkkqYxVct6ZyBIElK30RulTFHB3w8dzueKOHKIl5QRsRz6jBNVD0U96NOYi4gwSa/gLJOp49OgMToGdAje0BFkDnUFmQ+rOEp2SKp4XbPkQt13eYTbT2Kkx0Q51EHIIW6O14nebfKYH5ZawlUSQizzLBUsr2pJPF/BchhST2/puRzqq/Hz00p0DC2XKqh2F0s/tFBftVtHoXFMhHeSJxAbhYq7TwljZ9oOEAL8+hwhTiPUH+H+MZsURWKnWHLPQt8yUPXNgh5vuCMjrot7RK5V2GFIVJ1D51d1ylJiCzTDNKRFY9xuUURirxhOUxjFVRErTDS45nJoeegGYZKZHOlJ0jjA3kI6ijC5mc37PcQgCDNJAPUoRG3ZMbwv1We2jkEdG/np5AzvYKmkm21P8rwOw3aC6cJEYp7Z2LZjnqDxDibdue3QQM5pUuuLOY09dEVDE8YnXei1KEYMcRGyUKnk691dIxpxEexiNO+iEYVBSuGk39HF19qshjbdgLX16IYAUyARmS19n0SUrWyJ2S5ROuSNrmaiKlRWHKWzL+5Kqz4MjXKtXoUiymjTBo/6/o3Ya4ohyLMMRvYwGOX/aMgoZo6MogzaMrhM6YIwPTucD43g0hj6V4yPmFsRWTlQJ6wgQSTurlSANkZ7p5kiLct2Q3DNtmg5E2PI436/jiJBowoiVYuMSw/9VONsZCqchyxd/sYEe13L5eVlShBqm2+r+MpbmGvQGtUDu8FdAwskDn8YriR15XrF62ScwDqmVOgSlr3KL0wAdewNPfEa9nV+0QFjgajDXm5QKbymSpuBrx5sT2xBg1A9bBAccROBz4+a6qIEY6i+Nvrq+FmSXxyfF9ssi6jUS9MiQ5P84oMcfFzJy9x8n9seCQLATIg3DRjWKBhuby5KZfquCI0TfAg2VgQWZhtk8LYVtmvRvC2L1rJXbVmwln3flqm17Lgt851ChN/N7XwjOs922tkqKcVJm3kjWjPRDd9hwSWok5osbzX3m36W56fTSpT8Obe/xL8yUiw7op7H4O/W9tS6DhutFgRmFkUXmEpQl3nWpIooLho+TzwQxQg+loqWyfuKYpbEiK3eTVvuaxJ1jixZRbNN6eScT6usDGnnLabfIgJTZM0r2tPCoc8587o3PPZ45kxQj3XLCApGVhwaWquqEXGa99NQ2EGBRFLG4w28BRFrUAd/2WCWO3rZhb3qxl51YVfd2FUXdtGNXXRhVTdWddoLgnfDx8VCvgIVcq+xhi7BovK6iPlW39llbVUHCs7EDfBI33UDJRaLqxvkG3PbD6Uxr4N1RzfcJRJ3bNVac99jbwN81ACbhym4+nhrcQxS0R3jJnx10Vs213lXd7USUqz06ZDgebpeVVhZZyGVbNIgbqNBbNIgb6NBbtKgbqOhXdZUNFzdRsPfWhoYx7WKOCaiW/lxvSiGVjxacXn/FCvA4ru5omidTgP4ZdlfWkWSXNrftOewTARWXi3oHzqxtIW7oJ3AeRs47wL6pG1mkcM0ugW/amKvepS2gEXy6LS1hV1a9rCqv0XBYgqgaff5XreDfZ6KFnbWj121sas+bNN0xHZ7RNsgsSpgXYYM+6ZYXcXqo3fPmuMT9MHUTujD8728b8Q++qxKn3XSzfBJ7/DJbYbvo8+q9O7hDb3HX4m1a63JxnX3rLfg6hLJlEy4gwVtnV7Ue74At6hFHMCzaJYY4PKr/NQ9s3RJZptiDBB9LWoWQtopRs3eh9TgdW+jLj1Do2p2G1XmOttsXalRdKsUt1TZcps5b/UbeR8VFvruf9BC4z+j78lGffqyV5iHrX19eWTMfaybB/ryZPNAxuyNnvgxdotuw8X/ZTiG70sem2OYQ0TjsLkAgZ3lgXMOIrYmo/0oNQL9fr3s3TG9eAKt3OVVQnG44Ck+Ky3z1gtHYRAH+rypX4OFoM8gjWlrgRn63paFH1u/peIFssaqH1+xv3h3U7KI5yneMVa2Mx7t7WMWL3GaZy+SkMSKR1hYpQyyiT4dV0lrMKYGWR2rqDz1sxuP/KI8L30HLsOiSne/LntxLbHiwf8e6TFKj3ulHiVBnplrD+IIdDF4BH3yQ06xXjLXHoTLsXLXlx45njzyTF965LDEclOZt4/lewfHyV7kffDnR7l+K9EjDd08C0f20B11I77SLw2DiODjFb/toW5tAtJ4DcRWz5gCWKpd+PKkD8F4gKdvirZdt/oGffXmRZ7pSx9AiQgIWqUiE0LHdH5lfkqw4zjGwlC/nczGo8f7bpSv+xlZgZCQZNNapwNMdzbADhdegR6P6v1LGVJfYbdRU/QzWRl0eoMv+6F4JOox2iL9E0e7F6cc4D6qCK7Nb3UWb6Kr89KCxrwu7mxPmj/CtRtvp6PJeDT5y3j72cGg+Hw6+GLw+8GDwWTwePBs8MfB4eBk4A7Y4J+Dfw3+ffeTuw/vju9OC+jHH5Wc3w1qn7vf/A9mDaMZ</latexit> <latexit sha1_base64="A0803SRxkfT9iDkUxwV/3t5Rbms=">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</latexit> <latexit sha1_base64="A0803SRxkfT9iDkUxwV/3t5Rbms=">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</latexit> <latexit sha1_base64="wu6Hts1TOb3rsEaGlT7xgnbYqiA=">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</latexit> ✓j <latexit sha1_base64="I43qAb/N9+gEleDQ88LUjU6gBUM=">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</latexit> <latexit sha1_base64="caXfti+QmnIlgO/vw7RCZ7mPWrs=">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</latexit> <latexit sha1_base64="caXfti+QmnIlgO/vw7RCZ7mPWrs=">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</latexit> <latexit sha1_base64="isTL59a2OifKddQT14KzRD58m1M=">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</latexit> parameter latent observable augmented data approximate likelihood ratio Figure 2 A schematic of machine learning based approaches to likelihood-free inference in which the simulation provides training data for a neural network that is subsequently used as a surrogate for the intractable likelihood during inference. Reproduced from (Brehmer et al., 2018b). techniques (Brehmer et al., 2018c). In addition, an inference compilation technique has been applied to inference of a tau-lepton decay. This proof-of-concept eﬀort required developing probabilistic programming protocol that can be integrated into existing domain-speciﬁc simulation codes such as SHERPA and GEANT4 (Baydin et al., 2018; Casado et al., 2017). This approach provides Bayesian inference on the latent variables p(Z|X = x) and deep interpretability as the pos- the Hubble parameter evolution from type Ia supernova measurements. These experiences motivated the devel- opment of tools such as CosmoABC to streamline the ap- plication of the methodology in cosmological applica- tions (Ishida et al., 2015). More recently, likelihood-free inference methods based on machine learning have also been developed motivated by the experiences in cosmology. To confront the chal- lenges of ABC for high-dimensional observations X, a Brehmer+ 1805.00013

Physics Informed ML “although neural networks only work well for
an exponentially tiny fraction of all possible inputs, the laws of physics are such that the data sets we care about for machine learning are also drawn from an exponentially tiny fraction of all imaginable data sets…” “Why does deep and cheap learning work so well?” Lin, Tegmark, Rolnick arXiv:1608.08225 (2017)

• Computer vision: e.g., Spatial Transformer Network, GVNN (s03 layer
Euler,…) • High-energy physics: “QCD-Aware Recursive NN for Jet Physics” • Quantum Chemistry: “Ab-Initio Solution of the Many-Electron   Schrödinger Equation with Deep Neural Networks” • Louppe+ 1702.00748 Jaderberg+1506.02025; Handa+1607.07405 Pfau+ 1909.02487 Impart/impose/imbue physical constraints into architecture Euclidean Neural Networks rotation-, translation-, & permutation- equivariant convolutional neural networks for 3D point clouds for emulating ab initio calculations & generating atomic geometries Tess Smidt 2018 Alvarez Postdoctoral Fellow in Computing Sciences cf. "Machine learning and the physical sciences” Carleo+ 1903.10563

• Computer vision: e.g., Spatial Transformer Network, GVNN (s03 layer
Euler,…) • High-energy physics: “QCD-Aware Recursive NN for Jet Physics” • Quantum Chemistry: “Ab-Initio Solution of the Many-Electron   Schrödinger Equation with Deep Neural Networks” • Louppe+ 1702.00748 Jaderberg+1506.02025; Handa+1607.07405 Pfau+ 1909.02487 Impart/impose/imbue physical constraints into architecture Euclidean Neural Networks rotation-, translation-, & permutation- equivariant convolutional neural networks for 3D point clouds for emulating ab initio calculations & generating atomic geometries Tess Smidt 2018 Alvarez Postdoctoral Fellow in Computing Sciences Challenge: Find data embeddings & network architectures that conform to known taxonomies, conservation laws, & symmetries cf. "Machine learning and the physical sciences” Carleo+ 1903.10563

Raw Data Preprocessing Batch Predictions ɠ Featurization & Embedding Convert
raw data to features & embed following our understanding of the physical system Baking Physical Constraints into the Entire Learning Process ɡ Symmetry preserving layers ɢ Bottlenecks & Model Capacity Sparsity imposition ɣ Loss Function Curation Enforce physically meaningful instance-level predictions ɤ Distributional Loss Enforce ensemble-level predictions conform to expectations

VAE with Physical Parameters for Astronomical Time Series Input data
Light Curves ! + Label ! + Physical Params ! x l χ Reconstructed data Light Curve !x′ Latent Space x′ x l, χ Encoder q ϕ (z|x, l, χ) Decoder p θ (x|z, l, χ) z μ σ [l, χ] + labels + +… T eff With Jorge Martínez-Palomera (Berkeley)

Fast cosmic web simulations with generative adversarial networks Figure 2:
Comparison of summary statistics between N-body and GAN simulations, for box size of 500 Mpc. The statistics are: mass density histogram (upper left), peak count (upper right), power spectrum of 2D images (lower left) and cross power spectrum (lower right). The cross power spectrum is calculated between pairs N-body images (blue points), between pairs of GAN images (red points), and between pairs consisting of one GAN and one N-body image (cyan points). The power spectra are shown in units of h 1 Mpc, where h = H0/100 corresponds to the Hubble parameter. The standard errors on the mean of the shown with a shaded region, and are too small to be seen for the ﬁrst three panels. Rodriguez+ 1801.08070 Enabling Dark Energy Science with Deep Generative Models of Galaxy Images Ravanbaksh+16098.05769 5 . 7: Comparison of a C-VAE sample before and after adding noise a real COSMOS image with corresponding size, magnitude and shift. conditional models with increasing resolution in Denton al. (2015). In these conditional models, the generator : Z ⇥Y ! X and the discriminator D : X ⇥Y ! [0, 1], (a) Galaxy sizes (b) Galaxy brightness Fig. 8: Comparison of galaxy sizes and brightness between real COSMOS images and C-VAE samples. Colors indicate the value of the relevant variable used to condition the generated images (half-light radius for size and magnitude for brightness) accuracy and therefore the dynamics of this adversarial setting does not allow this mode of failure. Supernova (Thomas/Nugent); Exoplanets (Ford+11) Generative & Surrogate Modeling "Surrogate models for precessing binary black hole simulations with unequal masse” Varma+ 1905.09300 Idea: expensive simulations build on a coarse grid of input parameters are used to train a surrogate model to interpolate across parameter space Example: Numerical Relativity calculations of black hole merger waveforms

AI @ Scale Future

Big (GPU/TPU/Analog AI/…) Iron and the… Replicability Crisis Without massive-scale
access, only private companies will be able to develop & reproduce state of the art AI results

Big (GPU/TPU/Analog AI/…) Iron for…Optimization Neural Architecture Search Approaches Fixed
network Learn Weights Learn Hyperparameters Fixed Architecture Learn Architecture Fixed Task https://towardsdatascience.com/illustrated-eﬃcient-neural-architecture-search-5f7387f9fb6 ~PFlop-Hr SOTA ResNet-50 on ImageNet (Following Goyal et al. 1706.02677) ~100 PFLOP-Hr >10 ExaFlop-Hr?

‣Big Iron facilitates more experiments → p-hacking easier cf. Warden
(https://petewarden.com/2018/03/19/the-machine-learning-reproducibility-crisis/) The Darker Side of that Same Coin… → reproducibility crisis ‣Carbon footprint "Energy and Policy Considerations for Deep Learning in NLP” Strubell+ 1906.02243

Summary ‣ML is already central to astrophysical discovery & inference
at the scale ‣Novel questions pushing the envelope on computationally: prediction speed, memory consumption, etc. ‣Opportunity to Accelerate Learning (with less) on Physical Systems w/ Physics-based constraints ‣Growing symbiosis: ﬁrst-principles simulations ⟷ generative/surrogate/likelihood-free inference ‣Access to greater compute aids in replicability & optimization, but at a cost

Josh Bloom UC Berkeley (Astronomy), LBNL @profjsb ML at Scale:
Astrophysics AI for Science Town Hall 2019 Sept 11, LBNL Data Driven Discovery Investigator Thanks!

Machine Learning at Scale: Astrophysics

Machine Learning at Scale: Astrophysics

Joshua Bloom

More Decks by Joshua Bloom

Other Decks in Science

Featured

Transcript

Josh Bloom UC Berkeley (Astronomy), LBNL @profjsb ML at Scale:

E. Ramirez-Ruiz (UCSC) (cf. Kasen) 50 100 150 200 Days

Supernova Discovery in the Pinwheel Galaxy 11 hr after explosion

Discovery (& classification) on images is now a cottage industry

Cosmological Parameters from the Dark matter Distribution Figure 6. The

Challenge & Opportunity: Domain Adaptation Figure 1: Diagram of an

ML for Novel Discovery & Inference

https://www.forbes.com/sites/jillianscudder/2017/01/21/astroquizzical-care-planet-nine Searching for Planet 9 hypothesized to exist using mostly

M. Medford, Nugent, JSB et al. •More candidates than disk

Faster-than-GPU Inference is Already Being Explored in Particle Physics 1904.08986

See also Lang, Hogg, Schlegel 1410.7397, Leistedt+ 1807.01391 Forward-Folded Cosmology

Inverse Problems & Likelihood-Free Inference (a) DRDAE: quasi-circular (b) DRDAE:

Physics Informed ML “although neural networks only work well for

• Computer vision: e.g., Spatial Transformer Network, GVNN (s03 layer

• Computer vision: e.g., Spatial Transformer Network, GVNN (s03 layer

Raw Data Preprocessing Batch Predictions ɠ Featurization & Embedding Convert

VAE with Physical Parameters for Astronomical Time Series Input data

Fast cosmic web simulations with generative adversarial networks Figure 2:

AI @ Scale Future

Big (GPU/TPU/Analog AI/…) Iron and the… Replicability Crisis Without massive-scale

Big (GPU/TPU/Analog AI/…) Iron for…Optimization Neural Architecture Search Approaches Fixed

‣Big Iron facilitates more experiments → p-hacking easier cf. Warden

Summary ‣ML is already central to astrophysical discovery & inference

Josh Bloom UC Berkeley (Astronomy), LBNL @profjsb ML at Scale: