Sze-hon Kong

Array ( [id] => 15183257 [patent_doc_number] => 20190362220 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-11-28 [patent_title] => ATTENTIVE NEURAL COLLABORATIVE FILTERING FOR MODELING IMPLICIT FEEDBACK [patent_app_type] => utility [patent_app_number] => 15/989350 [patent_app_country] => US [patent_app_date] => 2018-05-25 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 9337 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -17 [patent_words_short_claim] => 197 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15989350 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14902393 [patent_doc_number] => 20190294962 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-09-26 [patent_title] => IMPUTATION USING A NEURAL NETWORK [patent_app_type] => utility [patent_app_number] => 15/973525 [patent_app_country] => US [patent_app_date] => 2018-05-07 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 16140 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -17 [patent_words_short_claim] => 304 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15973525 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14902387 [patent_doc_number] => 20190294959 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-09-26 [patent_title] => SCHEDULING NETWORK COMPUTATIONS [patent_app_type] => utility [patent_app_number] => 15/933225 [patent_app_country] => US [patent_app_date] => 2018-03-22 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 14169 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -17 [patent_words_short_claim] => 179 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15933225 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14283195 [patent_doc_number] => 20190138882 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-05-09 [patent_title] => METHOD AND APPARATUS FOR LEARNING LOW-PRECISION NEURAL NETWORK THAT COMBINES WEIGHT QUANTIZATION AND ACTIVATION QUANTIZATION [patent_app_type] => utility [patent_app_number] => 15/914229 [patent_app_country] => US [patent_app_date] => 2018-03-07 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 6032 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -20 [patent_words_short_claim] => 147 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15914229 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14751621 [patent_doc_number] => 20190258984 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-08-22 [patent_title] => GENERATIVE ADVERSARIAL NETWORKS IN PREDICTING SEQUENTIAL DATA [patent_app_type] => utility [patent_app_number] => 15/898847 [patent_app_country] => US [patent_app_date] => 2018-02-19 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 10623 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -17 [patent_words_short_claim] => 157 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15898847 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 18316912 [patent_doc_number] => 11630994 [patent_country] => US [patent_kind] => B2 [patent_issue_date] => 2023-04-18 [patent_title] => Optimized asynchronous training of neural networks using a distributed parameter server with eager updates [patent_app_type] => utility [patent_app_number] => 15/898433 [patent_app_country] => US [patent_app_date] => 2018-02-17 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 6 [patent_figures_cnt] => 6 [patent_no_of_words] => 8321 [patent_no_of_claims] => 21 [patent_no_of_ind_claims] => 3 [patent_words_short_claim] => 259 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => patent [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15898433 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14720733 [patent_doc_number] => 20190251430 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-08-15 [patent_title] => MIXED SIGNAL CMOS RPU WITH DIGITAL WEIGHT STORAGE [patent_app_type] => utility [patent_app_number] => 15/895162 [patent_app_country] => US [patent_app_date] => 2018-02-13 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 5805 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -17 [patent_words_short_claim] => 95 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15895162 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14689921 [patent_doc_number] => 20190244076 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-08-08 [patent_title] => Neural Network with Embedded Filter Layer [patent_app_type] => utility [patent_app_number] => 15/888812 [patent_app_country] => US [patent_app_date] => 2018-02-05 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 5777 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -18 [patent_words_short_claim] => 111 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15888812 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14658563 [patent_doc_number] => 20190236410 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-08-01 [patent_title] => BOOTSTRAPPING RECOMMENDATION SYSTEMS FROM PASSIVE DATA [patent_app_type] => utility [patent_app_number] => 15/886263 [patent_app_country] => US [patent_app_date] => 2018-02-01 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 5926 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -17 [patent_words_short_claim] => 92 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15886263 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14629475 [patent_doc_number] => 20190228105 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-07-25 [patent_title] => DYNAMIC WEBSITE CONTENT OPTIMIZATION [patent_app_type] => utility [patent_app_number] => 15/879351 [patent_app_country] => US [patent_app_date] => 2018-01-24 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 7760 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -17 [patent_words_short_claim] => 151 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15879351 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14629475 [patent_doc_number] => 20190228105 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-07-25 [patent_title] => DYNAMIC WEBSITE CONTENT OPTIMIZATION [patent_app_type] => utility [patent_app_number] => 15/879351 [patent_app_country] => US [patent_app_date] => 2018-01-24 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 7760 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -17 [patent_words_short_claim] => 151 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15879351 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14571727 [patent_doc_number] => 20190213470 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-07-11 [patent_title] => ZERO INJECTION FOR DISTRIBUTED DEEP LEARNING [patent_app_type] => utility [patent_app_number] => 15/865294 [patent_app_country] => US [patent_app_date] => 2018-01-09 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 5314 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -12 [patent_words_short_claim] => 64 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15865294 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14507509 [patent_doc_number] => 20190197409 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-06-27 [patent_title] => METHOD FOR UNSUPERVISED SEQUENCE LEARNING USING REINFORCEMENT LEARNING AND NEURAL NETWORKS [patent_app_type] => utility [patent_app_number] => 15/852512 [patent_app_country] => US [patent_app_date] => 2017-12-22 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 6907 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -17 [patent_words_short_claim] => 120 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15852512 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14507481 [patent_doc_number] => 20190197395 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-06-27 [patent_title] => MODEL ENSEMBLE GENERATION [patent_app_type] => utility [patent_app_number] => 15/851723 [patent_app_country] => US [patent_app_date] => 2017-12-21 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 4616 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -18 [patent_words_short_claim] => 124 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15851723 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14412267 [patent_doc_number] => 20190171977 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-06-06 [patent_title] => Using Machine Learning System to Dynamically Process Events [patent_app_type] => utility [patent_app_number] => 15/833049 [patent_app_country] => US [patent_app_date] => 2017-12-06 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 8592 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -21 [patent_words_short_claim] => 123 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15833049 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14379929 [patent_doc_number] => 20190163877 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-05-30 [patent_title] => DECISION SUPPORT FOR EFFECTIVE LONG-TERM DRUG THERAPY [patent_app_type] => utility [patent_app_number] => 15/822454 [patent_app_country] => US [patent_app_date] => 2017-11-27 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 8113 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -17 [patent_words_short_claim] => 46 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15822454 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 13611973 [patent_doc_number] => 20180357536 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2018-12-13 [patent_title] => SYSTEM FOR RECOGNIZING BEATING PATTERN AND METHOD FOR RECOGNIZING THE SAME [patent_app_type] => utility [patent_app_number] => 15/820927 [patent_app_country] => US [patent_app_date] => 2017-11-22 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 4027 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -18 [patent_words_short_claim] => 54 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15820927 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 13404229 [patent_doc_number] => 20180253657 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2018-09-06 [patent_title] => REAL-TIME CREDIT RISK MANAGEMENT SYSTEM [patent_app_type] => utility [patent_app_number] => 15/820340 [patent_app_country] => US [patent_app_date] => 2017-11-21 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 10345 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -16 [patent_words_short_claim] => 97 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15820340 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 14348505 [patent_doc_number] => 20190156225 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2019-05-23 [patent_title] => OPTIMIZING A HIERARCHICAL RULE-BASED DECISION POLICY [patent_app_type] => utility [patent_app_number] => 15/817396 [patent_app_country] => US [patent_app_date] => 2017-11-20 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 12121 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -17 [patent_words_short_claim] => 83 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15817396 [rel_patent_id] =>[rel_patent_doc_number] =>)

Array ( [id] => 13569831 [patent_doc_number] => 20180336463 [patent_country] => US [patent_kind] => A1 [patent_issue_date] => 2018-11-22 [patent_title] => SYSTEMS AND METHODS FOR DOMAIN-SPECIFIC OBSCURED DATA TRANSPORT [patent_app_type] => utility [patent_app_number] => 15/816899 [patent_app_country] => US [patent_app_date] => 2017-11-17 [patent_effective_date] => 0000-00-00 [patent_drawing_sheets_cnt] => 0 [patent_figures_cnt] => 0 [patent_no_of_words] => 13056 [patent_no_of_claims] => 0 [patent_no_of_ind_claims] => -15 [patent_words_short_claim] => 75 [patent_maintenance] => 1 [patent_no_of_assignments] => 0 [patent_current_assignee] =>[type] => publication [pdf_file] =>[firstpage_image] =>[orig_patent_app_number] => 15816899 [rel_patent_id] =>[rel_patent_doc_number] =>)