Pith. sign in

REVIEW 3 major objections 6 minor 39 references

A Novel Coded Computing Approach for Distributed Multi-Task Learning

T0 review · 3 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper claims a coded scheme that, under a data-placement condition, simultaneously attains the provable minimum uplink and downlink communication loads in distributed multi-task learning, including heterogeneous worker storage.

desk verdict The matrix-decomposition framing is fresh, but the main theorem is false: the lower bound only holds for N=K, and the construction permits negative d_k. read the letter →

arxiv 2507.18025 v1 pith:TUWEVFC5 submitted 2025-07-24 cs.IT math.IT

Minquan Cheng , Yongkang Wang , Lingyu Zhang , Youlong Wu This is my paper
classification cs.ITmath.IT MSC 94A1594B0568P30
keywords distributedmulti-tasklearningcodedcomputingcommunicationloadmatrixdecompositionMDScodesheterogeneousdataplacementlinearlyseparablecomputationHall'smarriagetheorem
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Distributed multi-task learning trains several personalized models at once, with workers sending local updates to a server and receiving a broadcast that lets each worker update its own model; communication, not computation, is usually the bottleneck. This paper tries to show that the communication bottleneck can be removed up to provable limits even when workers store unequal, arbitrary subsets of the data. It models the whole scheme as a matrix factorization and then constructs one code that simultaneously achieves the smallest possible uplink and downlink traffic under a mild data-placement condition. A sympathetic reader should care because earlier optimal coded schemes only worked in symmetric settings; if the construction is right, optimal communication becomes available in general heterogeneous deployments, and the same design transfers to distributed linearly separable computation.

What carries the argument

The load-bearing object is the matrix decomposition $P=AB$, which unifies the two communication directions. The uplink matrix $P$ stacks each worker's coded transmissions; its zero-column pattern reflects exactly which packets worker $k$ cannot compute. The downlink matrix $B$ is an MDS matrix over a finite field with $q > \max\{S, N(K-1)\}$, so all its square submatrices are invertible, and the server broadcasts $B'\mathbf{v}$. The invertible matrix $A$ is assembled from row blocks $A_k$ that annihilate the columns of $B$ corresponding to packets worker $k$ lacks. Condition 1, $|\cap_{i\in[k]} P_i| \le S - \sum_{i\in[k]} d_i$ for every $k\in[K]$, is the Hall-type combinatorial condition that guarantees such an $A$ exists: for every prefix of workers, the intersection of their missing-packet sets is small enough that the zero blocks can be matched disjointly. Hall's Marriage Theorem turns this inequality into the disjoint subsets needed to prove the row blocks are linearly independent.

What would settle it

A concrete disproof would be a placement with $N\neq K$ that satisfies Condition 1 yet has $d_k = N + (K-1)|Z_k| - \sum_{k'} |Z_{k'}|$ negative for some worker, or has the claimed $L^*_{\text{up}}$ smaller than $L^*_{\text{down}}$; either observation would make the scheme's per-worker packet counts and the server's broadcast impossible. One candidate to check is $K=4,N=6$ with $Z_1=\{1\}$ and $Z_2=Z_3=Z_4=\{1,\ldots,6\}$, where the formula gives $L^*_{\text{up}}=5/3$, $L^*_{\text{down}}=5$, and $d_1=-10$.

Watch

Extended reading notes

Core claim

The paper's central claim is Theorem 1: in a $(K,N;\{Z_k\})$ DMTL system satisfying Condition 1, there exists a scheme with minimum communication loads $L^*_{\text{up}} = \frac{KN}{K-1} - \sum_{z\in[K]} \frac{z|I_z|}{K-1}$ and $L^*_{\text{down}} = N - \min_k |Z_k|$, where $I_z$ is the set of data batches stored by exactly $z$ workers. The construction splits every local update into $K-1$ packets and represents transmissions by matrices: the uplink matrix $P$ has zero blocks exactly where a worker lacks data, and the downlink matrix $B$ is chosen with the MDS (Maximum Distance Separable) property, so that any sufficiently small square submatrix is invertible. The proof shows that Condition 1, an intersection bound on the sets of packets absent from each worker, is exactly what is needed for the row blocks $A_1,\ldots,A_K$ to span the full space, making $A$ invertible in $P=AB$. If correct, this is the first optimal coded MTL scheme that handles heterogeneous data placement, and the uplink half of the same construction is claimed to solve distributed linearly separable computation for arbitrary placement.

Load-bearing premise

Everything rests on the lower-bound formula for the uplink load being true for every $N$, although the cited lemma proves it only for $N=K$; when $N>K$ the formula can come out smaller than the downlink load, which would make the claimed optimum impossible even if Condition 1 holds.

Editorial extensions

If this is right

  • For any placement satisfying Condition 1, the proposed scheme's loads equal $L^*_{\text{up}}$ and $L^*_{\text{down}}$, so no coded scheme can do better for those placements.
  • The optimality subsumes the symmetric equal-load cases treated by earlier schemes and extends to unequal computation loads and asymmetric data placement.
  • The downlink broadcast size is $N - \min_k |Z_k|$ regardless of total storage, so the least-stored worker determines the broadcast cost.
  • The uplink construction transfers to distributed linearly separable computation, where it handles non-cyclic heterogeneous data placements that prior linearly separable computation schemes could not.
  • In the paper's $(K=4,N=6)$ numerical comparison, the scheme matches the lower bounds at every tested computation load while the prior schemes leave an uplink gap.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The matrix-decomposition view suggests a natural research direction: formulating the converse bound itself as a rank constraint on the uplink matrix $P$, which would give a unified proof for all $N$ and all placements.
  • Condition 1 is checkable directly from the storage sets, so practitioners could add a quick feasibility test before deploying the scheme; a natural next step is an efficient algorithm that verifies it and outputs the matrices $A$ and $B$.
  • Because $B$ is an MDS matrix, a straggler-tolerant variant suggests itself: if a worker fails to upload its $d_k$ packets, the server may still broadcast as long as the remaining rows of $P$ span enough of the space, and verifying when erased rows preserve invertibility is a direct extension.
  • The same decomposition may apply to decentralized settings where no central server exists, since the broadcast step only needs a common MDS matrix available to all workers; whether the invertibility condition still holds under gossip-style communication is untested.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper considers a distributed multi-task learning (DMTL) system with K workers and N data batches, where each worker stores an arbitrary subset Zk. It proposes to model the uplink and downlink encoding as a matrix factorization P = AB, with B an MDS matrix, and claims that under a placement-dependent condition (Condition 1) the scheme achieves the information-theoretically optimal uplink and downlink communication loads for arbitrary heterogeneous placements. The proof is based on a lower bound cited from [12], a Hall-marriage argument for constructing an invertible A, and a decoding step using the MDS property. The paper also reports numerical comparisons with [12] and [33] and claims an extension to distributed linearly separable computation.

Significance. If correct, the result would be a substantial step: previous optimal coded MTL schemes were limited to symmetric placements, and a general optimal scheme for heterogeneous placements has been an open problem. The matrix-decomposition formulation is a useful unifying viewpoint, and the paper contains a fully worked example and numerical comparisons. However, the central optimality claim is not valid as stated. The lower bound used for the uplink is taken from a lemma stated only for N = K and is false for a concrete (K = 4, N = 6) instance; moreover, Condition 1 does not prevent the prescribed number of transmitted packets d_k from being negative, so the proposed algorithm is not even executable for some instances satisfying its hypotheses. A third issue in the proof of Lemma 6 (the Hall step) is demonstrably false for a positive-d_k instance. These are load-bearing, not cosmetic, defects. The conceptual framework may still be salvageable, but the paper in its current form does not establish the advertised optimality.

major comments (3)
  1. [Section III.A, Theorem 1 with Lemma 1] The lower bound L*_up is imported from Lemma 1, which is explicitly restricted to (K, N = K) systems. Theorem 1 then asserts the same expression for arbitrary N, and Proposition 1 uses it. This extension is not only unproved but false. Consider K = 4, N = 6, Z1 = {1}, Z2 = Z3 = Z4 = {1,...,6}. Here d1 = -10, d2 = d3 = d4 = 5, S = 5; the claimed L*_up is (24 - 19)/3 = 5/3, while the claimed L*_down is 6 - 1 = 5. In any scheme, the server's broadcast X is a deterministic function of the uplink messages, so the downlink normalized load cannot exceed the total uplink normalized load; hence the pair (5/3, 5) cannot be achieved. The claimed formula for L*_up is therefore invalid for N != K, and Theorem 1's optimality statement collapses for the general setting advertised in the paper.
  2. [Section III.A, Eq. (11), Proposition 1, and Algorithm 1] The construction requires each worker k to transmit d_k coded packets, but Condition 1 does not imply d_k >= 0. In the same (K = 4, N = 6) instance as above, d1 = -10 while Condition 1 holds: |P1| = 15 = S - d1, P2 = P3 = P4 = empty, so all inequalities in Condition 1 reduce to 0 <= S - sum_i d_i, which hold with equality at k = 4. Algorithm 1 line 15 would need a matrix A1 with -10 rows, and Proposition 1 asks worker 1 to send -10 packets. Thus Theorem 1 asserts existence of a scheme for a system in which its own scheme cannot be executed; the theorem is internally inconsistent as stated. The examples in the paper all have positive d_k, but that is not guaranteed by the stated hypotheses.
  3. [Appendix C, Lemma 6 and the inequality |Pk ∩ Pi| <= S - (dk + di)] This inequality is used to ensure |Ai| >= di, but it does not follow from Condition 1 and is false in general. Take K = 3, N = 6, Z1 = {5,6}, Z2 = {1,2,3,4}, Z3 = {3,4,6}. Then d = (1,5,3), S = 9, P1 = {1,2,3,4,7,8,9,10}, P2 = {5,6,11,12}, P3 = {1,2,5,7,8,11}. Condition 1 holds: |P1| = 8 <= 8, |P1 ∩ P2| = 0 <= 3, and the triple intersection is empty. Yet |P2 ∩ P3| = 2 > S - (d2 + d3) = 1. Lemma 6 would require a subset of P3 of size d2 = 5 disjoint from P2, but P3 \ P2 has only 4 elements, so the claimed matching does not exist. Since Lemma 6 underlies Lemma 5 and hence Lemma 4, the Hall-marriage proof of Theorem 1 is not valid even for instances with all d_k positive.
minor comments (6)
  1. [Algorithm 1 header] Algorithm 1 is titled '(K, N=K; ...)' but is applied to general (K,N); the header should be corrected.
  2. [Lemmas 2 and 3] Lemmas 2 and 3 state that the subsets {Z_k} are subsets of [K]; for the general (K,N) model they should be subsets of [N].
  3. [Section III.C example] The Vandermonde example contains corrupted entries such as 'w w 11' and 'w w 14'; these should be cleaned up.
  4. [Section IV.A, matrix P_{k,j} definition] The definition of P_{k,j} says 'dk > 0 rows'; since d_k can be zero or negative under the stated assumptions, the definition should either impose d_k >= 0 as an explicit hypothesis or be reformulated.
  5. [Remark 3, LSC extension] Remark 3's LSC extension is dimensionally unclear: B' has (K-1)(N - min_k r_k) rows, while the matrix F of the LSC problem has N_c rows; the claimed identification needs a separate argument.
  6. [Reference [34]] Reference [34] is cited as 'Hall's marriage theorem, 2025'; a standard textbook or Hall's original paper would be more appropriate.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular reduction found: the scheme is an explicit achievability construction whose loads match the cited [12] lower bound by design and whose correctness rests on internal MDS/Hall lemmas; the theorem fails instead by applying the N=K-only bound out of scope and by allowing d_k<0 under Condition 1.

full rationale

The derivation chain is not circular in any of the defined senses. Theorem 1's claimed optimal values are matched by construction: equation (11) sets d_k = N + (K-1)r_k - sum r_k' so that S = sum d_k = KN - sum_z z|I_z| equals (K-1) times the claimed L*_up, and lambda = (K-1)(N - min_k r_k) yields the claimed L*_down; this bound-matching is the standard, non-circular structure of an achievability proof. The real substantive content is the existence of the invertible matrix A in P = AB, which is proved internally through Lemma 4, Lemma 5, Lemma 6, the MDS property of B, and Hall's Marriage Theorem, and none of those lemmas presupposes the optimal communication-load values; no parameter is fitted to any data. The converse half of the optimality claim is imported from Lemma 1, which cites [12] and overlaps with present author Y. Wu. However, [12] is an externally published, parameter-free result within its stated assumption, and the paper itself writes the lemma as 'Given a (K, N = K; ...) DMTL scheme', so the citation is real evidence for exactly that scope and does not, by itself, make the reasoning circular. What the paper then does is silently reuse the same formula for N different from K in Theorem 1 and in Section IV-A ('From (8) we have S >= (K-1)L*_up'), where the bound is unproven and in fact false: for K=4, N=6, Z1={1}, Z2=Z3=Z4={1,...,6}, the formula gives S = 5 and L*_up = 5/3, which is less than the downlink load 5 and hence impossible. Relatedly, Condition 1 holds in that same instance yet (11) yields d1 = -10, so the Section IV construction of d_k x S matrices A_k is undefined, and Algorithm 1 itself is labeled '(K, N=K; ...)' inconsistently with Theorem 1. These are soundness and correctness defects of an over-generalized theorem, not reductions of the conclusion to the hypotheses; no fitted input is renamed as a prediction and no known result is relabeled. Score 1 reflects only the mildly self-referential origin of the converse bound, which retains independent content within its published N=K scope.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central claim rests on the cited lower bound from [12] (applied beyond its proven range), on the nonnegativity of the derived packet counts, and on Condition 1, which is introduced specifically to make the invertibility proof work. No fitted parameters or invented physical entities appear.

assumptions (5)
  • ad hoc to paper The lower bounds L*_up and L*_down from [12] apply to the general (K,N) setting.
    Invoked in Proposition 1 and Theorem 1, but Lemma 1 only proves them for N=K. For N≠K the uplink bound is false, as shown by a counterexample.
  • domain assumption Each worker k can send d_k coded packets with d_k = N + (K-1) r_k - R.
    Used in Proposition 1 and Algorithm 1. Requires d_k ≥ 0, which is not guaranteed under Condition 1.
  • standard math MDS matrices exist over F_q for q > max{S, N(K-1)}.
    Standard coding theory result, used to construct the downlink matrix B.
  • ad hoc to paper Condition 1 is satisfied for the considered placements.
    The scheme's invertibility proof depends on Condition 1. It is not shown to hold broadly, and its statement is insufficient to justify the pairwise bounds used in Lemma 6.
  • domain assumption The global update is non-linear but the communication scheme is linear and each worker decodes all local updates.
    Model assumption from [12]. The workers recover all local update values and then compute the non-linear global update locally.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Novel Coded Computing Approach for Distributed Multi-Task Learning." pith.science (2026). https://pith.science/paper/TUWEVFC5

@misc{pith2026250718025,
  author       = {Pith},
  title        = {Pith review of: A Novel Coded Computing Approach for Distributed Multi-Task Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TUWEVFC5}},
  note         = {Machine review of arXiv:2507.18025}
}
read the original abstract

Distributed multi-task learning (DMTL) effectively improves model generalization performance through the collaborative training of multiple related models. However, in large-scale learning scenarios, communication bottlenecks severely limit practical system performance. In this paper, we investigate the communication bottleneck within a typical DMTL system that employs non-linear global updates. This system involves distributed workers, assisted by a central server, who collaboratively learn distinct models derived from a non-linear aggregation of their local model parameters. We first characterize the communication process as a matrix decomposition problem. It transforms workers' data storage constraints into structural characteristics of the uplink encoding matrix, and worker data retrieval demands into Maximum Distance Separable (MDS) properties of the downlink encoding matrix. Building on this, we propose a novel coded DTML scheme that can greatly reduce the communication cost of the DTML with heterogeneous data placement. Theoretical analysis demonstrates that the proposed scheme achieves the theoretical lower bound for communication overhead under mild conditions. Remarkably, this optimality holds for both traditional homogeneous computing environments and various heterogeneous scenarios. Furthermore, our scheme is extensible to a distributed linearly separable computation problem where the target function involves multiple linear combinations of local update values. This indicates that our scheme offers a new way of tackling heterogeneous data placement challenges in various distributed applications.

Figures

Figures reproduced from arXiv: 2507.18025 by the authors.

Figure 1
Figure 1. A DMTL system consists of a central server connected to [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The (K = 4, N = 6, {Zk | k ∈ [4]}) DMTL scheme with data placement in Table I As illustrated in [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

39 extracted references · 38 canonical work pages

  1. [12]

    Communication-efficient coded computing for distributed multi-task learning,

    H. Hu, Y . Wu, Y . Shi, S. Li, C. Jiang, and W. Zhang, “Communication-efficient coded computing for distributed multi-task learning,” IEEE Transactions on Communications , 2023

    work page 2023

  2. [33]

    On exploiting network topology for hierarchical coded multi-task learning,

    H. Hu, S. Li, M. Cheng, S. Ma, Y . Shi, and Y . Wu, “On exploiting network topology for hierarchical coded multi-task learning,” IEEE Transactions on Communications, vol. 72, no. 8, pp. 4930–4944, 2024

    work page 2024

  3. [1]

    Communication-efficient edge ai: Algorithms and systems,

    Y . Shi, K. Yang, T. Jiang, J. Zhang, and K. B. Letaief, “Communication-efficient edge ai: Algorithms and systems,” IEEE Communications Surveys & Tutorials, vol. 22, no. 4, pp. 2167–2191, 2020

    work page 2020

  4. [2]

    Edge artificial intelligence for 6g: Vision, enabling technologies, and applications,

    K. B. Letaief, Y . Shi, J. Lu, and J. Lu, “Edge artificial intelligence for 6g: Vision, enabling technologies, and applications,” Selected Areas in Communications, IEEE Journal on (J-SAC) , vol. 40, no. 1, 2022

    work page 2022

  5. [3]

    A fundamental tradeoff between computation and communication in distributed computing,

    S. Li, M. A. Maddah-Ali, Q. Yu, and A. S. Avestimehr, “A fundamental tradeoff between computation and communication in distributed computing,” IEEE Transactions on Information Theory , vol. 64, no. 1, pp. 109–128, 2018

    work page 2018

  6. [4]

    Fundamental limits of caching,

    M. A. Maddah-Ali and U. Niesen, “Fundamental limits of caching,” IEEE Transactions on Information Theory , vol. 60, no. 5, pp. 2856–2867, 2014

    work page 2014

  7. [5]

    Mapreduce: simplified data processing on large clusters,

    J. Dean and S. Ghemawat, “Mapreduce: simplified data processing on large clusters,” Commun. ACM, vol. 51, no. 1, p. 107–113, 2008

    work page 2008

  8. [6]

    A scalable framework for wireless distributed computing,

    S. Li, Q. Yu, M. A. Maddah-Ali, and A. S. Avestimehr, “A scalable framework for wireless distributed computing,” IEEE/ACM Transactions on Networking, vol. 25, no. 5, pp. 2643–2654, 2017

    work page 2017

Show all 39 references
  1. [7]

    Wireless mapreduce distributed computing,

    F. Li, J. Chen, and Z. Wang, “Wireless mapreduce distributed computing,” IEEE Transactions on Information Theory , vol. 65, no. 10, pp. 6101–6114, 2019

    2019

  2. [8]

    Dof of a cooperative x-channel with an application to distributed computing,

    Y . Bi, P. Ciblat, M. Wigger, and Y . Wu, “Dof of a cooperative x-channel with an application to distributed computing,” in 2022 IEEE International Symposium on Information Theory (ISIT) . IEEE, 2022, pp. 566–571

    2022

  3. [9]

    Coded computing for half-duplex wireless distributed computing systems via interference alignment,

    Y . Wu, Z. Huang, K. Yuan, S. Ma, and Y . Bi, “Coded computing for half-duplex wireless distributed computing systems via interference alignment,” arXiv preprint arXiv:2310.15598 , 2023

    2023 arXiv

  4. [10]

    Coded computing for master-aided distributed computing systems,

    H. Chen and Y . Wu, “Coded computing for master-aided distributed computing systems,” in 2020 IEEE Information Theory Workshop (ITW) . IEEE, 2021, pp. 1–5

    2020

  5. [11]

    On the optimality of data exchange for master-aided edge computing systems,

    H. Chen, J. Long, S. Ma, M. Tang, and Y . Wu, “On the optimality of data exchange for master-aided edge computing systems,” IEEE Transactions on Communications, vol. 71, no. 3, pp. 1364–1376, 2023

    2023

  6. [13]

    Coded distributed computing for hierarchical multi-task learning,

    H. Hu, S. Li, M. Cheng, and Y . Wu, “Coded distributed computing for hierarchical multi-task learning,” in 2023 IEEE Information Theory Workshop (ITW). IEEE, 2023, pp. 480–485

    2023

  7. [14]

    A new combinatorial coded design for heterogeneous distributed computing,

    N. Woolsey, R.-R. Chen, and M. Ji, “A new combinatorial coded design for heterogeneous distributed computing,” IEEE Transactions on Communications, 2021

    2021

  8. [15]

    New results on the computation-communication tradeoff for heterogeneous coded distributed computing,

    F. Xu, S. Shao, and M. Tao, “New results on the computation-communication tradeoff for heterogeneous coded distributed computing,” IEEE Transactions on Communications , vol. 69, no. 4, pp. 2254–2270, 2021

    2021

  9. [16]

    Coded computation over heterogeneous clusters,

    A. Reisizadeh, S. Prakash, R. Pedarsani, and A. S. Avestimehr, “Coded computation over heterogeneous clusters,” IEEE Transactions on Information Theory, vol. 65, no. 7, pp. 4227–4242, 2019

    2019

  10. [17]

    Coded distributed computing with pre-set assignments of data and output functions,

    Y . Wang and Y . Wu, “Coded distributed computing with pre-set assignments of data and output functions,” arXiv preprint arXiv:2201.06300 , 2022

    2022 arXiv

  11. [18]

    Joint design of shuffling and function assignment in heterogeneous coded distributed computing,

    H. Song, K. Son, and W. Choi, “Joint design of shuffling and function assignment in heterogeneous coded distributed computing,” IEEE Transactions on Signal Processing , vol. 70, pp. 2560–2575, 2022

    2022

  12. [19]

    Cascaded coded distributed computing schemes based on placement delivery arrays,

    J. Jiang and L. Qu, “Cascaded coded distributed computing schemes based on placement delivery arrays,” IEEE Access , vol. 8, pp. 221 385–221 395, 2020

    2020

  13. [20]

    Asymptotically optimal cascaded coded distributed computing via combinatorial designs,

    M. Cheng, Y . Wu, and X. Li, “Asymptotically optimal cascaded coded distributed computing via combinatorial designs,” arXiv preprint arXiv:2302.05826, 2023

    2023 arXiv

  14. [21]

    A combinatorial design for cascaded coded distributed computing on general networks,

    N. Woolsey, R.-R. Chen, and M. Ji, “A combinatorial design for cascaded coded distributed computing on general networks,” IEEE Transactions on Communications, vol. 69, no. 9, pp. 5686–5700, 2021

    2021

  15. [22]

    Communication efficient distributed machine learning with the parameter server,

    M. Li, D. G. Andersen, A. Smola, and K. Yu, “Communication efficient distributed machine learning with the parameter server,” in Advances in Neural Information Processing Systems , Z. Ghahramani, M. Welling, C. Cortes, N. Lawrence, and K. Weinberger, Eds., vol. 27. Curran Asso...

  16. [23]

    Gradient coding: avoiding stragglers in distributed learning,

    R. Tandon, Q. Lei, A. G. Dimakis, and N. Karampatziakis, “Gradient coding: avoiding stragglers in distributed learning,” in Proceedings of the 34th International Conference on Machine Learning , vol. 70, 2017, p. 3368–3376

    2017

  17. [24]

    Communication-computation efficient gradient coding,

    M. Ye and E. Abbe, “Communication-computation efficient gradient coding,” 2018

    2018

  18. [25]

    Adaptive gradient coding,

    H. Cao, Q. Yan, X. Tang, and G. Han, “Adaptive gradient coding,” Networking, IEEE/ACM Transactions on , vol. 30, no. 2, 2022

    2022

  19. [26]

    Distributed linearly separable computation,

    K. Wan, H. Sun, M. Ji, and G. Caire, “Distributed linearly separable computation,” IEEE Transactions on Information Theory , vol. 68, no. 2, pp. 1259–1278, 2022

    2022

  20. [27]

    Fundamental limits of distributed linearly separable computation under cyclic assignment,

    W. Huang, K. Wan, H. Sun, M. Ji, R. C. Qiu, and G. Caire, “Fundamental limits of distributed linearly separable computation under cyclic assignment,” in 2023 IEEE International Symposium on Information Theory (ISIT) , 2023, pp. 2296–2301

    2023

  21. [28]

    Communication-efficient coded distributed multi-task learning,

    H. Tang, H. Hu, K. Yuan, and Y . Wu, “Communication-efficient coded distributed multi-task learning,” in 2021 IEEE Global Communications Conference (GLOBECOM), 2021, pp. 1–6

    2021

  22. [29]

    A survey on multi-task learning,

    Y . Zhang and Q. Yang, “A survey on multi-task learning,” IEEE Transactions on Knowledge and Data Engineering , 2021

    2021

  23. [30]

    Distributed multi-task relationship learning,

    S. Liu, S. J. Pan, and Q. Ho, “Distributed multi-task relationship learning,” in Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , 2017, pp. 937–946

    2017

  24. [31]

    Federated multi-task learning,

    V . Smith, C.-K. Chiang, M. Sanjabi, and A. Talwalkar, “Federated multi-task learning,” in Proceedings of the 31st International Conference on Neural Information Processing Systems , 2017, pp. 4427–4437

    2017

  25. [32]

    Feature hashing for large scale multitask learning,

    K. Q. Weinberger, A. Dasgupta, J. Attenberg, J. Langford, and A. J. Smola, “Feature hashing for large scale multitask learning,” CoRR, vol. abs/0902.2206, 2009. [Online]. Available: http://arxiv.org/abs/0902.2206

    2009 arXiv

  26. [34]

    Hall’s marriage theorem,

    P. J. Cameron, “Hall’s marriage theorem,” 2025

    2025

  27. [35]

    End-to-end multi-task learning with attention,

    S. Liu, E. Johns, and A. J. Davison, “End-to-end multi-task learning with attention,” in 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2019, pp. 1871–1880

    2019

  28. [36]

    Xing and S

    C. Xing and S. Ling, Coding Theory: A First Course . Coding Theory: A First Course, 2003

    2003

  29. [37]

    F. J. Macwilliams and N. J. A. Sloane, The Theory of Error-Correcting Codes . The Theory of Error-Correcting Codes, 1977

    1977

  30. [38]

    On mds codes via cauchy matrices,

    R. M. Roth and A. Lempel, “On mds codes via cauchy matrices,” IEEE Transactions on Information Theory , vol. 35, no. 6, pp. 1314–1319, 1989

    1989

  31. [2014]

    Available: https://proceedings.neurips.cc/paper_files/paper/2014/file/935ad074f32d1e8f085a143449894cdc-Paper.pdf

    [Online]. Available: https://proceedings.neurips.cc/paper_files/paper/2014/file/935ad074f32d1e8f085a143449894cdc-Paper.pdf

    2014

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.