byteshift

8 ms·

1.
Downsample Parquet Files(github.com)
1 points by byteshift 5y ago|0 comments
2.
by byteshift 6y ago
We index geospatial data using a learned index in this work (cf. Section 3): http://cidrdb.org/cidr2021/papers/cidr2021_paper19.pdf Code: https://github.com/learnedsystems/RadixSpline
3.
by byteshift 6y ago
For a detailed study of learned indexes, see this work: https://vldb.org/pvldb/vol14/p1-marcus.pdf All code is available in open source: https://github.com/learnedsystems/SOSD
4.
SOSD: A Benchmark for Learned Indexes(github.com)
3 points by byteshift 6y ago|0 comments
5.
RadixSpline: A Single-Pass Learned Index(github.com)
47 points by byteshift 6y ago|3 comments
6.
Cuckoo Index: A Lightweight Secondary Index Structure(github.com)
2 points by byteshift 6y ago|0 comments
7.
by byteshift 6y ago
My understanding is that this is even feasible without location information if you use an approximate filter (e.g., a Cuckoo Filter [1]), which they propose in DP-3T (II). With 20 bits/key such filters yield a false positive rate (FPR)
8.
Who Wins? Machine Learning or Traditional Indexes(learned.systems)
4 points by byteshift 7y ago|0 comments
9.
Estimating the Number of Results of Database Queries with Deep Learning(arxiv.org)
3 points by byteshift 8y ago|0 comments
10.
by byteshift 9y ago
For this year's contest, the task is to evaluate batches of database join queries on a set of pre-defined relations. The challenge is to execute the queries as fast as possible without (much) prior indexing. So the time has finally com
11.
Show HN: 2018 ACM SIGMOD Programming Contest – $10K Prizes Donated by Microsoft(sigmod18contest.db.in.tum.de)
13 points by byteshift 9y ago|3 comments
12.
Approximate Geospatial Joins with Precision Guarantees [pdf](db.in.tum.de)
3 points by byteshift 9y ago|0 comments
13.
by byteshift 9y ago
Cool stuff! We actually built a similar system to analyze the NYC taxi dataset (or any other geospatial dataset). We use the (PostgreSQL wire protocol compliant) HyPerSpace database in the backend [1]. Let us know if you're ever experi