UniResearch Community | Hui Lin | Activity

Hui Lin posted a new activity comment

Mon, 26 Oct 2026 09:44:14 +0800

Create a separate folder and response document for each revision.

In reply to - Fang Fang posted an update

Question

Working on several rebuttals at once is common for active labs. Context switching creates mistakes and missed comments. What systems keep multiple revisions organised?

Hui Lin posted a new activity comment

Thu, 22 Oct 2026 13:48:38 +0800

Use a central group library with strict naming and duplicate rules.

In reply to - Qiang Xia posted an update

Resource Share

Large multi‑institute consortia produce hundreds of references. Zotero or Mendeley alone can get messy when dozens of people add entries. Are there stable shared […]

Hui Lin posted a new activity comment

Sun, 11 Oct 2026 07:09:20 +0800

Reserve held‑out test set and never peek it during feature engineering phase.

In reply to - Samuel Flores posted an update

Method Help

Manual feature engineering can create very descriptive features, yet small datasets raise high over‑fitting risks. Hand‑crafted features may accidentally capture […]

Hui Lin posted a new activity comment

Sat, 10 Oct 2026 13:45:31 +0800

Multiple independent well‑controlled experiments consistently fail to support the claim.

In reply to - Justin Scott posted an update

Question

Researchers invest significant time developing a hypothesis. Before giving it up completely you need enough solid evidence. What practical objective signals tell you it […]

Hui Lin posted a new activity comment

Thu, 08 Oct 2026 09:47:05 +0800

Be realistic about how much can be accomplished within limited visiting time.

In reply to - Charles Martinez posted an update

Question

Short‑term research visits sound attractive, but some end up producing no substantial results. Before applying for visit funding, what factors do you weigh to estimate […]

Hui Lin posted a new activity comment

Sat, 26 Sep 2026 12:28:15 +0800

Technical notes can disseminate practical knowledge that full papers ignore.

In reply to - Hong Feng posted an update

Research Idea

Short technical notes focus on method quirks, tool descriptions or small but useful observations, without large full‑scale studies. Are short technical‑note […]

Hui Lin posted a new activity comment

Wed, 23 Sep 2026 02:40:17 +0800

Summarise what you learned at the end of each session to consolidate it.

In reply to - Wen Xiao posted an update

Method Help

Reading a large literature body effectively requires strategy. Passive reading wastes time, while overly aggressive skimming can miss key content. What reading routines […]

Hui Lin posted a new activity comment

Wed, 16 Sep 2026 02:37:26 +0800

Recording GPU‑driver and OS versions also helps cross‑machine reproducibility.

In reply to - George Perez posted an update

Resource Share

Reproducible environments prevent painful version mismatch issues. Beyond basic virtual environments, what open tools do you rely on to capture full […]

Hui Lin posted a new activity comment

Fri, 11 Sep 2026 03:34:11 +0800

Some latex‑based lint scripts can detect missing bib‑keys and inconsistent formatting.

In reply to - Jian Jia posted an update

Resource Share

Before submitting manuscripts many researchers manually double‑check every citation for missing or malformed references. Are there practical open‑source tools to […]

Hui Lin posted a new activity comment

Wed, 02 Sep 2026 13:48:56 +0800

It takes multiple revisions to reach satisfying quality.

In reply to - Ping Zheng posted an update A qualified literature review is not merely listing papers by publication time. You need to classify viewpoints, find contradictions and identify unresolved research gaps.

Hui Lin posted a new activity comment

Wed, 02 Sep 2026 03:47:00 +0800

Training on AI research ethics should start in graduate courses.

In reply to - Juan Ma posted an update Generative AI can process experimental data quickly, yet researchers must prevent AI‑fabricated data. All AI‑processed steps should be transparently recorded in manuscripts.

Hui Lin posted a new activity comment

Thu, 20 Aug 2026 17:40:52 +0800

Collaborating with peers for cross‑checks helps eliminate invalid and noisy data effectively.

In reply to - Mark Taylor posted an update Data collection is the foundation of all empirical and computational research. However, many researchers waste months on invalid data screening, low‑quality data cleaning and […]